TWI632360B - Pod screening system and pod screening device - Google Patents
Pod screening system and pod screening device Download PDFInfo
- Publication number
- TWI632360B TWI632360B TW105104823A TW105104823A TWI632360B TW I632360 B TWI632360 B TW I632360B TW 105104823 A TW105104823 A TW 105104823A TW 105104823 A TW105104823 A TW 105104823A TW I632360 B TWI632360 B TW I632360B
- Authority
- TW
- Taiwan
- Prior art keywords
- image
- pod
- width
- inspection
- screening system
- Prior art date
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/34—Sorting according to other particular properties
- B07C5/342—Sorting according to other particular properties according to optical properties, e.g. colour
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/85—Investigating moving fluids or granular solids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Immunology (AREA)
- Geometry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Pathology (AREA)
- Analytical Chemistry (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Sorting Of Articles (AREA)
- Image Processing (AREA)
- Length Measuring Devices By Optical Means (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
[課題] 提供一種豆莢篩選系統及豆莢篩選裝置,係觀察殼內的豆子時,減去利用透射光的拍攝需求,可利用拍攝反射光來檢查殼內豆子的容納狀態(尤其是豆粒數及大小)。 [解決手段] 一種豆莢篩選系統,係藉由分析拍攝帶殼豆莢的影像,對前述豆莢進行檢查或篩選,具備:影像取得手段,係取得攝影手段所拍攝的影像;以及影像分析手段,係對影像取得手段所取得的影像進行分析;前述影像分析手段係執行:尺寸檢查,係根據攝影影像測量各個豆莢長度與寬度;以及豆粒數檢查,係根據所測量豆莢長度來計算殼中容納的豆粒數。[Question] A pod screening system and a pod screening device are provided. When observing the beans in the shell, the need to use transmitted light is subtracted, and the reflected light can be used to check the accommodation state of the beans in the shell (especially the number and size of the beans). ). [Solutions] A pod screening system detects or screens the pods by analyzing the images of the shell pods, and has: image acquisition means, images obtained by photographing means; and image analysis means, The image obtained by the image acquisition means is analyzed; the image analysis means performs: size inspection, measuring the length and width of each pod according to the photographic image; and checking the number of beans, based on the measured length of the pod to calculate the number of beans contained in the shell .
Description
本發明係關於一種豆莢篩選系統以及豆莢篩選裝置,係以毛豆為首的豆莢,尤其是篩選帶殼的豆莢,亦關於一種根據外觀及/或植被指數(vegetation index)篩選豆莢用之豆莢篩選系統以及豆莢篩選裝置。The present invention relates to a pod screening system and a pod screening device, which are pods headed by edamame, especially for screening shelled pods, and also for a pod screening system for screening pods according to appearance and/or vegetation index. Pod screening device.
豆莢為常見於豆科植物之果實,包含:菜豆(kidney bean)、豌豆(garden pea)、合歡(albizia julibrissin)、大豆、紫荊(cercis chinensis)、羽扇豆(lupine)、三葉草(clover)、野豌豆(vetch)等。這些豆莢大多在帶殼狀態下收成,在收成未熟成大豆之毛豆中,大多以帶殼狀態供至餐桌。Bean pods are common in legumes and include: kidney bean, garden pea, albizia julibrissin, soybean, cercis chinensis, lupine, clover, wild Peas (vetch) and so on. Most of these pods are harvested in a shelled state, and most of the edamame that is not matured into soybeans is supplied to the table in a shelled state.
使上述豆莢流通至市面時,為了管理其品質或等級需要進行篩選。然而,豆莢為各個殼內的豆子數量皆不同,且該豆子大小亦不同。因此,以往提出數個應對帶殼的豆莢的篩選進行自動化,並利用攝影機,篩選以毛豆為首的豆莢之技術。When the pods are circulated to the market, screening is required in order to manage their quality or grade. However, the number of beans in each shell is different, and the size of the beans is also different. Therefore, in the past, several techniques for responding to the screening of shelled pods have been proposed, and a technique of screening pods headed by edamame has been screened using a camera.
例如專利文獻1(日本特開2008-62116號公報)中提出一種豆莢篩選系統,具備:透射光攝影手段,係利用透射光並根據豆莢等級篩選豆莢,藉由透射豆莢的透射光對豆莢進行攝影;根據該透射光攝影手段所拍攝的豆莢影像來判定豆莢等級並進行篩選。又,專利文獻2(日本特開2005-279524號公報)中提出一種毛豆莢精密篩選方法,係為了達成毛豆莢的精密篩選作業的省工、高效率,在相對鏡(opposite mirrors)的連接部的中心面上,使毛豆莢的長邊方向呈垂直狀態下掉落來取得毛豆莢的整面影像,該相對鏡係於平面觀視V字型上以120度角連接二面鏡子。For example, Patent Document 1 (JP-A-2008-62116) proposes a pod screening system including: a transmitted light photographing method that uses transmitted light to screen pods according to the pod level, and photographs the pods by transmitting light of the transmissive pods. According to the image of the pod image taken by the transmitted light photographing means, the pod level is determined and screened. Further, Patent Document 2 (JP-A-2005-279524) proposes a method for accurately screening edamame pods, which is a joint portion of opposite mirrors in order to achieve labor saving and high efficiency of precision screening operations of edama pods. On the center surface, the long side direction of the bean pod is dropped vertically to obtain a full-face image of the bean pod, and the relative mirror is connected to the two mirrors at a 120-degree angle on the V-shape of the plane view.
又,豆莢類的篩選時,較理想為進一步對豆莢的活性度(activity)、含水量等之成長狀況進行篩選。接著,以往亦提出一種對拍攝作物的影像進行頻譜分析(spectral analysis),判斷成長狀況等技術。Further, in the screening of the pods, it is preferable to further filter the growth status of the pods, such as the activity and the water content. Next, a technique for performing spectrum analysis on the image of crops and judging growth conditions has been proposed.
例如專利文獻3(日本特開2006-250827號公報)中提出一種作物的成長狀況分析方法,係藉由利用遙控感測(remote sensing)之頻譜(spectrum)影像來推測稻米的蛋白質含量從而診斷食用品質。又,專利文獻4(日本特開2013-231645號公報)中提出一種茶葉摘採適合性評價方法,係為了以短時間且高精密度對茶樹的新芽是否摘採簡單地進行判斷,使用含有茶葉影像資訊的光學資料來計算植被指數,並使用該植被指數對茶葉的摘採適合性進行評價。For example, Patent Document 3 (JP-A-2006-250827) proposes a method for analyzing the growth state of a crop by using a spectrum image of remote sensing to estimate the protein content of rice for diagnosis. quality. Further, a method for evaluating the suitability of tea picking and extracting is proposed in the patent document 4 (JP-A-2013-231645), in order to judge whether or not the new shoot of the tea tree is picked up in a short time and with high precision, and to use the tea image information. The optical data is used to calculate the vegetation index, and the vegetation index is used to evaluate the suitability of the tea.
[先前技術文獻] [專利文獻] 專利文獻1:日本特開2008-62116號公報 專利文獻2:日本特開2005-279524號公報 專利文獻3:日本特開2006-250827號公報 專利文獻4:日本特開2013-231645號公報[Prior Art Document] [Patent Document 1] Japanese Patent Laid-Open Publication No. Hei. No. Hei. No. Hei. No. Hei. JP-A-2013-231645
[發明所欲解決之課題] 如上所述,以往利用攝影機篩選帶殼的豆莢之技術已提出各種提案。然而,為了檢查殼內豆子的容納狀態,必須拍攝該豆莢的透射光。接著,為了拍攝豆莢中的透射光,需要有如以網眼狀皮帶進行搬送或拍攝掉落豆莢等之拍攝透射光用之構成。[Problems to be Solved by the Invention] As described above, various proposals have been made in the art for screening a shelled pod by a camera. However, in order to check the accommodation state of the beans in the shell, it is necessary to photograph the transmitted light of the pod. Next, in order to photograph the transmitted light in the pod, it is necessary to carry out the photographing of the transmitted light such as the mesh belt or the photographing of the pod.
因此,本發明第1課題係提供一種減去如上所述的透射光拍攝需求,利用拍攝照射在帶殼豆莢上光的反射光,可檢查殼內豆子的容納狀態(尤其是豆粒數或大小)之豆莢篩選系統及豆莢篩選裝置。Therefore, the first object of the present invention is to provide a method of reducing the amount of transmitted light that is irradiated onto a shell bean pod by subtracting the need for transmitted light imaging as described above, and checking the state of accommodation of the beans in the shell (especially the number or size of beans). The pod screening system and the pod screening device.
又,第2課題係提供一種在檢查殼內的豆子容納狀態,同時檢查豆莢的大小、形狀、殼的裂痕、以及缺口等之情形時,使用拍攝反射光的影像來減少品質管理用影像的儲存區域,並提升檢查處理速度之豆莢篩選系統及豆莢篩選裝置。In addition, the second problem is to reduce the size and shape of the pod, the crack of the shell, and the gap when the bean is inspected in the inspection case, and the image of the reflected light is used to reduce the storage of the image for quality control. Area, and improve the processing speed of the pod screening system and pod screening device.
又,以往豆莢的篩選、檢查所使用的攝影機為可對可見光進行拍攝的攝影機,僅藉由獨特的可見光影像只對豆莢外觀等進行檢查。此種拍攝可見光之影像檢查係因為利用獨特平面影像中的頻譜、色彩的變化等進行判斷,因此其正確性仍存在待改善之處。尤其是在檢查豆莢的刮痕等之情形時,單就可見光影像仍難以明確顯現該刮痕部的顏色與周圍顏色之差異,因此難以正確檢查、篩選。Moreover, the camera used for the screening and inspection of the pods is a camera capable of capturing visible light, and only the appearance of the pods and the like are inspected only by the unique visible light image. Such an image inspection for capturing visible light is judged by using a spectrum, a color change, or the like in a unique planar image, and thus the correctness still needs to be improved. In particular, when the scratches of the pods are examined, it is difficult to clearly show the difference between the color of the scratched portion and the surrounding color in the visible light image alone, and thus it is difficult to accurately inspect and screen.
因此,本發明第3課題係提供一種不拘泥於可見光影像,可更正確地執行外觀的檢查、篩選,尤其是等同目視般檢查、篩選刮痕等之豆莢篩選系統、豆莢篩選裝置、以及豆莢篩選方法。Therefore, the third object of the present invention is to provide a pod screening system, a pod screening device, and a pod screening which are more suitable for visual inspection and screening, and particularly for visual inspection, screening of scratches, and the like, without being restricted to visible light images. method.
又,屬於植物的豆莢會因成長狀況、收成時期等生長速度會有所不同。又,收成後亦會隨時間、保管狀況,豆莢的乾燥度(含水量)會有所不同。然而,將豆莢作為商品出貨時,必須符合一定的品質基準。因此,除了品質管理徹底進行,仍必須檢查且篩選豆莢的活性度。In addition, the pods belonging to plants will vary in growth rate due to growth conditions and harvest periods. In addition, the dryness (water content) of the pods will vary depending on the time and storage conditions after the harvest. However, when pods are shipped as commodities, they must meet certain quality standards. Therefore, in addition to thorough management of quality, it is still necessary to check and screen the activity of the pods.
因此,本發明第4課題係提供一種利用植被指數,進一步對豆莢的活性度、含水量、氮量、纖維量、蛋白質含量中至少一個進行檢查或篩選之豆莢篩選系統、豆莢篩選裝置、以及豆莢篩選方法。Therefore, the fourth object of the present invention is to provide a pod screening system, a pod screening device, and a pod for further inspecting or screening at least one of pod activity, water content, nitrogen amount, fiber amount, and protein content using a vegetation index. Screening method.
[用以解決課題之手段] 為了解決上述課題,本發明係提供一種豆莢篩選系統以及豆莢篩選裝置,係藉由分析所取得攝影影像,可判斷殼內的豆子數量,進一步根據攝影影像亦可判斷有無缺口、蒂頭。[Means for Solving the Problems] In order to solve the above problems, the present invention provides a pod screening system and a pod screening device, which are capable of determining the number of beans in a shell by analyzing the obtained photographed image, and further determining the number of beans according to the photographed image. There are no gaps, pedicels.
亦即,本發明係提供一種豆莢篩選系統,係藉由分析拍攝帶殼豆莢的影像,對前述豆莢進行檢查或篩選,具備:影像取得手段,係取得攝影手段所拍攝的影像;以及影像分析手段,係對影像取得手段所取得的影像進行分析;前述影像分析手段係執行:尺寸檢查,係根據攝影影像測量各個豆莢長度與寬度;以及豆粒數檢查,係根據所測量豆莢的長度來計算殼中容納的豆粒數。That is, the present invention provides a pod screening system for inspecting or screening the pods by analyzing and photographing the shell pods, and having: image acquisition means, images obtained by photographing means; and image analysis means The image obtained by the image acquisition means is analyzed; the image analysis means performs: size inspection, measuring the length and width of each pod according to the photographic image; and checking the number of beans, calculating the shell according to the length of the measured pod The number of beans contained.
又,為了解決前述課題中之至少一課題,本發明係提供一種豆莢篩選系統,係藉由分析拍攝帶殼豆莢的影像,對前述豆莢進行檢查或篩選,具備:影像取得手段,係取得攝影手段拍攝豆莢的反射影像;以及影像分析手段,分析影像取得手段所取得的影像;前述影像取得手段係對由紅色波長區域的反射光所構成的紅色影像、以及由近紅外光波長區域的反射光所構成的近紅外光影像進行抽出或取得;前述影像分析手段係利用下列公式1計算植被指數,並藉由前述植被指數的數值進行檢查或篩選豆莢。 [公式1][公式中 R:紅色影像中紅色波長區域的反射率 IR:近紅外光影像中近紅外光波長區域的反射率]Moreover, in order to solve at least one of the problems, the present invention provides a pod screening system for inspecting or screening the pods by analyzing and photographing the image of the shell pods, and having the means for obtaining images and obtaining photographing means a reflection image of the pod; and an image analysis means for analyzing the image obtained by the image acquisition means; the image acquisition means is a red image composed of the reflected light of the red wavelength region and a reflected light of the wavelength region of the near-infrared light The constructed near-infrared light image is extracted or obtained; the image analysis means calculates the vegetation index by the following formula 1, and checks or screens the pod by the value of the aforementioned vegetation index. [Formula 1] [R in the formula: reflectance of red wavelength region in red image IR: reflectance of near-infrared light wavelength region in near-infrared light image]
豆莢篩選系統可利用電腦來構成。具體而言,利用電腦軟體,以控制處理內容與處理步驟之電腦來具體實現。接著,前述影像分析手段係藉由執行電腦軟體來進行處理,亦由CPU或記憶體所構成。The pod screening system can be constructed using a computer. Specifically, the computer software is used to control the computer that processes the content and the processing steps. Then, the image analysis means is processed by executing a computer software, and is also composed of a CPU or a memory.
<豆粒數檢查> 在本發明之豆莢篩選系統中,影像分析手段係執行「尺寸檢查」,該尺寸檢查係從取得的攝影影像取得有關各個豆莢的長度與寬度的資訊。關於該長度與寬度的資訊除了根據所取得的攝影影像的畫素來取得作為點數(或畫素)之外,可取得檢查對象物實際的長度與寬度。當從攝影影像計算檢查對象物的實際長度及寬度時,對攝影影像上被特定之長度乘上固定放大或縮小的常數。<Bean Grain Number Inspection> In the pod screening system of the present invention, the image analysis means performs "size inspection" which acquires information on the length and width of each pod from the acquired photographic image. The information on the length and the width can be obtained as the number of points (or pixels) from the pixels of the captured image, and the actual length and width of the object to be inspected can be obtained. When the actual length and width of the inspection object are calculated from the photographic image, the specific length on the photographic image is multiplied by a constant that is fixedly enlarged or reduced.
接著,根據所測量豆莢的長度來計算容納於殼中的豆子數量。該豆子數量的設定係對豆莢的長度賦予關連,並根據定義豆子數量的資料進行設定。例如準備預先對豆莢的長度與豆子的數量賦予關連之長度/數字表(Number table),根據影像的分析,將讀取的豆莢長度作為搜尋值進行搜尋,藉此可計算、抽出對應前述長度之豆子的數量。又,根據取得的豆莢長度,可以輸出固定的數值(豆子的數量)之方式設計程式。亦即,在本發明中,藉由測量殼的長度並使用該數值,不需拍攝透射光就可計算容納於殼內的豆子的大致數量。Next, the number of beans contained in the shell is calculated based on the length of the measured pod. The setting of the number of beans is related to the length of the pods, and is set according to the data defining the number of beans. For example, it is prepared to preliminarily associate the length of the pod with the number of beans/Number table, and according to the analysis of the image, the length of the read pod is searched as a search value, thereby calculating and extracting the corresponding length. The number of beans. Further, according to the length of the obtained pod, the program can be designed in such a manner that a fixed value (the number of beans) can be output. That is, in the present invention, by measuring the length of the shell and using the value, the approximate number of beans contained in the shell can be calculated without photographing the transmitted light.
接著,上述豆粒數檢查時,計算殼內豆子的數量,同時計算因存在的豆子數量而會改變的殼寬度。亦即,前述影像分析手段中豆粒數檢查係在長度方向上分割該豆粒數的數量、各個豆莢的影像。接著,在分割的各個區域中寬度最大的位置設為凸部,同時在該凸部間之寬度比凹部還更窄且為最小的位置設為凹部。接著,以計算該凸部與凹部寬度之方式構成。如此一來,藉由取得殼的長度、凸部及凹部的寬度,將其數值與預設的容許值進行比較,可更詳細地檢查作為檢查對象之豆莢的尺寸。Next, when the number of beans is checked, the number of beans in the shell is calculated, and the width of the shell which changes due to the number of beans present is calculated. In other words, in the image analysis means, the number of beans is divided into the number of the number of beans and the image of each pod in the longitudinal direction. Next, the position where the width is the largest among the divided regions is a convex portion, and the position where the width between the convex portions is narrower and smaller than the concave portion is a concave portion. Next, it is configured to calculate the width of the convex portion and the concave portion. In this way, by taking the length of the shell, the width of the convex portion and the concave portion, and comparing the value with the preset allowable value, the size of the pod as the inspection object can be inspected in more detail.
又,在設定凸部及凹部,並計算其寬度時,不用在長度方向上分割各個豆莢的影像,仍可特定凸部及凹部並計算其寬度。亦即,在前述豆莢篩選系統中,影像分析手段係執行:尺寸檢查,係從攝影影像測量各個豆莢長度與寬度;以及豆粒數檢查,係根據所測量豆莢的長度設定容納於殼中豆子數量,同時根據豆莢的長度及/或豆子的數量設定寬度方向中凸部的數量,並取得設定凸部之各個寬度、凸部彼此之間作為最小寬度之凹部的寬度。Further, when the convex portion and the concave portion are set and the width thereof is calculated, it is not necessary to divide the image of each pod in the longitudinal direction, and the convex portion and the concave portion can be specified and the width thereof can be calculated. That is, in the pod screening system described above, the image analysis means performs: size inspection, measuring the length and width of each pod from the photographic image; and checking the number of beans, and setting the number of beans contained in the shell according to the length of the measured pod. At the same time, the number of the convex portions in the width direction is set in accordance with the length of the pods and/or the number of beans, and the widths of the concave portions which define the respective widths of the convex portions and the convex portions as the minimum width are obtained.
藉由執行前述尺寸檢查,來計算豆莢的長度及/或豆子的數量,其後根據前述豆莢的長度及/或種子的數量設定寬度方向中凸部的數量。在此,根據豆莢的長度及/或豆子的數量進行設定之數量為凸部的數量,並非特定該凸部位置。該凸部的位置係掃瞄豆莢全長的寬度,僅對依寬度數值由大往小之順序設定凸部數量進行抽出,將該抽出的寬度位置設為凸部,並計算該凸部中的各個寬度。例如當殼內的豆子數量設為2顆時,將該寬度方向中的凸部數量設定為2,依寬度較廣之順序將2處設為凸部,並計算其寬度。又,殼內的豆子數量設為3顆時,將該寬度方向中凸部的數量設為3,依寬度較廣之順序將3處設為凸部,並分別取得凸部的寬度。The length of the pod and/or the number of beans are calculated by performing the aforementioned dimensional inspection, and thereafter the number of the convex portions in the width direction is set according to the length of the pod and/or the number of seeds. Here, the number set according to the length of the pod and/or the number of beans is the number of the convex portions, and the convex portion position is not specified. The position of the convex portion is the width of the entire length of the scanned pod, and only the number of convex portions is set in order from the width to the small value, and the extracted width position is set as a convex portion, and each of the convex portions is calculated. width. For example, when the number of beans in the shell is two, the number of convex portions in the width direction is set to 2, and two portions are set as convex portions in the order of wide width, and the width thereof is calculated. When the number of beans in the shell is three, the number of the convex portions in the width direction is set to three, and three portions are formed as convex portions in the order of wide width, and the width of the convex portion is obtained.
其次,在凸部彼此之間,抽出殼的寬度為最小之位置,將該最小寬度位置設為凹部,並取得其寬度。亦即,殼內存在有豆子時,可從輪廓膨起來設定凸部。接著,因豆子之間存在有凹部,因此藉由確認凹部的存在,可確定凸部存在豆子。Next, between the convex portions, the width of the shell is minimized, and the minimum width position is set as a concave portion, and the width thereof is obtained. That is, when there is a bean in the shell, the convex portion can be set from the outline. Then, since there is a concave portion between the beans, it is possible to confirm the presence of the beans in the convex portion by confirming the existence of the concave portion.
又,前述豆粒數檢查中,凸部以2個以上進行計算時,較理想為確認該凸部間之寬度的變化量。僅利用殼的長度來特定豆粒數時,即便殼長一樣,仍依據殼內存在的豆子大小而定,可能為1顆亦可能為2顆。因此,較理想為凸部間之寬度的變化若出現山谷部(凹部)則可確認寬度發生變化。In the above-described inspection of the number of beans, when the convex portions are calculated by two or more, it is preferable to confirm the amount of change in the width between the convex portions. When only the length of the shell is used to specify the number of beans, even if the shell length is the same, depending on the size of the beans present in the shell, there may be two or two. Therefore, it is preferable to change the width of the convex portion so as to change the width when the valley portion (concave portion) appears.
在豆粒數檢查中,較理想為特定容納於殼內的豆子僅為1顆的豆莢。豆子為1顆之豆莢首先利用殼的全長進行特定。亦即,可設想較基準長度還短的豆莢的豆子僅為1顆。根據豆莢的長度預測豆粒數,藉此可提升豆粒數檢查中處理速度。然而,對於具有基準值以上長度之豆莢,將判斷殼內的豆子為2顆以上。然而,即便判斷為2顆以上,但豆子較大時,實際上亦有僅容納1顆之情形。因此,從攝影影像讀取寬度,抽出、確認凸部與凹部,並計算實際的豆粒數。亦即,為基準值以上的長度且豆粒數設為2時,在長邊方法將該豆莢劃分呈二個區域,在各個區域對寬度最的凸部進行抽出。其結果,若容納於殼內的豆子為2顆以上,則該凸部會有2處,該凸部間必定存在凹部。然而,殼內的豆子為1顆時,假設凸部設為2處,但該凸部彼此之間並未存在比二個凸部還窄的寬度,故並未存在凹部。因此,對殼內的豆子可能為1顆之豆莢,可確認凸部間是否存在凹部,便可判斷是否為1顆豆子。In the inspection of the number of beans, it is preferable that only one bean is contained in the shell. The bean pods are first made using the full length of the shell. That is, it is conceivable that only one bean has a bean pod that is shorter than the reference length. The number of beans is predicted based on the length of the pods, thereby increasing the processing speed in the number of beans. However, for pods having a length greater than the reference value, it is judged that the number of beans in the shell is two or more. However, even if it is judged to be two or more, when the beans are large, there is actually a case where only one is accommodated. Therefore, the width is read from the photographic image, and the convex portion and the concave portion are extracted and confirmed, and the actual number of beans is calculated. In other words, when the length is equal to or greater than the reference value and the number of beans is set to 2, the pod is divided into two regions by the long side method, and the convex portion having the most width is extracted in each region. As a result, when two or more beans are contained in the case, there are two convex portions, and a concave portion is necessarily present between the convex portions. However, when the number of beans in the shell is one, it is assumed that the convex portions are two, but the convex portions are not narrower than the two convex portions, so that the concave portions are not present. Therefore, it is possible to have one bean pod in the shell, and it can be confirmed whether there is a concave portion between the convex portions, and it can be judged whether it is one bean.
如此,特定殼內的豆子為1顆豆莢的處理特別有助篩選毛豆時。在毛豆的篩選中,從殼內的豆子數量亦會影響商品價值時,豆子可特定1顆之豆莢,藉由篩選此設定即可提高商品的價值。In this way, the treatment of a bean pod in a specific shell is particularly useful for screening edamame. In the screening of edamame, when the number of beans in the shell also affects the value of the commodity, the bean can be specified for one pod, and by screening this setting, the value of the commodity can be increased.
藉此,根據本發明之豆莢篩選系統,無需拍攝透射光,僅分析從攝影手段所取得的攝影影像,即可正確掌握容納於殼內的豆子數量。Thereby, according to the pod screening system of the present invention, it is possible to accurately grasp the number of beans contained in the shell without analyzing the transmitted light and analyzing only the photographed image obtained by the photographing means.
如上所述分析攝影影像時,較理想為對含有複數個豆莢之攝影影像進行各莢果抽出處理。因此,前述影像分析手段係較理想為,對從攝影手段取得的攝影影像進行標記處理,對每個豆莢創建影像。進一步,較理想為,對各影像進行角度修正,對經角度修正後的修正影像測量作為長軸方向的長度、與作為短軸方向的寬度。藉由執行標記處理,可特定豆莢的輪廓形狀,又,藉由進行角度修正,可更正確地進行豆莢的長度及寬度測定。When analyzing a photographic image as described above, it is preferable to perform a pod extraction process on a photographic image containing a plurality of pods. Therefore, it is preferable that the image analysis means performs marking processing on the captured image obtained by the photographing means, and creates an image for each pod. Further, it is preferable that the angle correction is performed on each of the images, and the corrected image after the angle correction is measured as the length in the long axis direction and the width in the short axis direction. By performing the labeling process, the contour shape of the pod can be specified, and by performing the angle correction, the length and width of the pod can be more accurately measured.
亦即,豆莢的篩選進行自動化時,必需以皮帶輸送機等搬送手段搬送,並拍攝搬送的豆莢,並必須對搬送的豆莢進行檢查。然而,使由搬送手段所搬送的豆莢方向正確地對齊是相當困難的。因此,拍攝輕易搬送而來的豆莢,將其方向進行修正,藉此可實現一邊符合檢查的實際情況一邊正確地進行豆莢長度及寬度的測定之豆莢篩選系統。In other words, when the screening of the pods is automated, it is necessary to carry them by means of a conveyance means such as a belt conveyor, and to photograph the transferred pods, and it is necessary to inspect the transferred pods. However, it is quite difficult to properly align the direction of the pods carried by the transport means. Therefore, the pods that are easily transported are photographed, and the direction is corrected, whereby the pod screening system that accurately measures the length and width of the pods while meeting the actual conditions of the inspection can be realized.
<殼的缺口檢查> 又,在本發明之豆莢篩選系統中,較理想為,前述影像分析手段對作為檢查對象物的豆莢檢查是否有缺口。 亦即,前述影像分析手段係較理想為,對於從前述攝影手段取得的攝影影像,將寬度變化量為既定範圍的區域設定作為檢查區域,取得該檢查區域中任意的基準位置的寬度並作為基準寬度,同時取得從前述基準位置朝固定方向移動後的位置之寬度並作為比較寬度,將算出基準寬度與比較寬度之差分與容許值進行比較,並進行缺口檢查。In the pod screening system of the present invention, it is preferable that the image analysis means check whether or not the pod as the inspection object is notched. In other words, the image analysis means preferably sets an area in which the width change amount is a predetermined range as an inspection area for the captured image obtained by the imaging means, and obtains a width of an arbitrary reference position in the inspection area as a reference. The width is simultaneously obtained as the width of the position moved from the reference position toward the fixed direction, and the difference between the reference width and the comparative width is compared with the allowable value, and the notch inspection is performed.
藉由將寬度變化量為既定範圍的區域設定作為檢查區域,可除去長度方向的端部等、容易頻生干擾的區域(亦即為易辯識區域),藉此可提升解析速度。接著,在該檢查區域中,取得任意訂定基準位置的寬度並作為基準寬度,並取得從該基準位置朝固定方向移動後的位置之寬度並作為比較寬度,將算出基準寬度與比較寬度之差分與容許值進行比較,並進行缺口檢查。亦即,在發生缺口部分,因豆莢的寬度大幅變化,對變化量較大的豆莢,可判斷發生缺口。By setting the region in which the amount of change in width is within the predetermined range as the inspection region, it is possible to remove the region (such as an easily identifiable region) which is likely to cause frequent interference, such as the end portion in the longitudinal direction, thereby improving the analysis speed. Then, in the inspection region, the width of the arbitrary reference position is obtained as the reference width, and the width of the position moved from the reference position in the fixed direction is obtained as the comparison width, and the difference between the reference width and the comparison width is calculated. Compare with the allowable value and perform a gap check. That is, in the case where the notch is generated, the width of the pod is greatly changed, and the pod having a large amount of change can be judged to have a gap.
又,在檢查區域中取得基準寬度之基準位置亦可為任意位置,例如可將前述凸部與凹部之任一個設為基準值。在凸部設為基準位置時,從該基準位置朝固定方向移動之位置的寬度係至凹部為止會漸漸地變窄。另一方面,在凹部設為基準位置時,從該基準位置朝固定方向移動之位置的寬度係至凸部為止會漸漸地變寬。藉此,利用此種豆莢類形狀,可判斷比較寬度的變化量增減,可更加高速且正確地檢查有無發生缺口。Further, the reference position at which the reference width is obtained in the inspection region may be an arbitrary position. For example, any one of the convex portion and the concave portion may be set as a reference value. When the convex portion is set as the reference position, the width of the position moving from the reference position toward the fixed direction gradually becomes narrower until the concave portion. On the other hand, when the concave portion is set as the reference position, the width of the position moved from the reference position in the fixed direction gradually increases until the convex portion. Thereby, by using such a pod shape, it is possible to judge whether the amount of change in the comparative width is increased or decreased, and it is possible to check the presence or absence of a notch more quickly and accurately.
<殼的蒂頭檢查> 又,在本發明之豆莢篩選系統中,較理想為前述影像分析手段係進行蒂頭檢查,該蒂頭檢查係進一步調查從殼細長延伸的有無蒂頭。 亦即,前述影像分析手段係進一步對從前述攝影手段取得的攝影影像進行二值化處理;在設定為殼的長度方向之任意複數地點中,較理想為,檢測寬度方向中變色的邊緣的數量,對該邊緣的數量為3個以上的地點特定作為蒂頭存在部位,較理想為,進一步從前述蒂頭存在部位的起點與終點計算前述蒂頭的長度。<Ticket inspection of the shell> Further, in the pod screening system of the present invention, it is preferable that the image analysis means perform a pedicle inspection, and the pedicle inspection further investigates the presence or absence of a pedicle extending from the shell. In other words, the image analyzing means further performs binarization processing on the imaged image obtained from the image capturing means; and in any of a plurality of places set in the longitudinal direction of the shell, it is preferable to detect the number of edges which are discolored in the width direction. The position where the number of the edges is three or more is specified as the position where the pedicle is present, and it is preferable that the length of the pedicle is further calculated from the start point and the end point of the pedicle presence portion.
對攝影影像執行二值化處理,藉此攝影影像中豆莢的輪廓會變清晰,可輕易區分背景部分與豆莢部分。其結果,豆莢部分的區域變得容易辯識,除了謀求提升處理速度之外,可正確地辯識影像,也可提升檢查精密度。Binary processing is performed on the photographic image, whereby the outline of the pod in the photographic image becomes clear, and the background portion and the pod portion can be easily distinguished. As a result, the area of the pod portion is easily recognized, and in addition to improving the processing speed, the image can be correctly recognized, and the inspection precision can be improved.
作為該蒂頭檢查的檢查對象為從殼的邊緣部分呈細長線條狀(stripe shape)突出之細長突起部等,利用影像辯識來判讀該細長突起部,並依據其長度進行篩選。具體而言,在創建每個豆莢的影像中,利用寬度方向中作為輪廓部分的邊緣的數量進行判斷。亦即,遍及豆莢的長度方向,將朝寬度方向延伸的線與輪廓部分之交點設為邊緣,該邊緣的數量若為2,則利用殼的輪廓來取得,邊緣的數量若超過2時,判斷殼的輪廓存在有突出的部分(亦即為蒂頭)。接著,將該處理遍及豆莢的長度方向的整體上執行,若邊緣的數量超出2時則計算從存在有突起部至該邊緣的數量為2為止的部分之區域長度並設為蒂頭的長度。The object to be inspected for the pedicle inspection is an elongated protrusion or the like which protrudes from the edge portion of the shell in a stripe shape, and the elongated protrusion is interpreted by image recognition, and is screened according to the length thereof. Specifically, in the image in which each pod is created, the number of edges in the width direction as the contour portion is judged. That is, the intersection of the line extending in the width direction and the contour portion is set as an edge throughout the length direction of the pod, and if the number of the edge is 2, the contour of the shell is used, and if the number of edges exceeds 2, it is judged The outline of the shell has a protruding portion (that is, a pedicle). Next, the processing is performed over the entire length of the pod, and when the number of edges exceeds 2, the length of the region from the portion where the projection is present to the edge is 2 is calculated and the length of the tip is set.
其結果,判讀拍攝反射光之2維影像,僅進行分析即可檢查蒂頭的存在。在此觀點上,若暫時拍攝透射光的影像,則會因光的強度可能使該蒂頭無法被拍攝,因此為了對從攝影影像檢查有無該蒂頭,拍攝反射光是能達成其目的。As a result, a two-dimensional image of the reflected light is read, and the presence of the pedicle can be checked only by analysis. From this point of view, if the image of the transmitted light is temporarily captured, the pedicle may not be photographed due to the intensity of the light. Therefore, in order to check the presence or absence of the pedicle from the photographic image, it is possible to obtain the reflected light.
<豆莢的活性度、含水量、氮量、纖維量、蛋白質含量檢查> 該檢查係利用植被指數。亦即,植被指數(尤其是常態化差異植被指數)為利用植物表面的反射光來判斷植物的植被狀況或活性度用之指標。亦即,包含於種植物的葉綠素在0.5~0.7μm的紅色可見光區域之反射率未滿20%,相對於此在0.7~1.3μm的近紅外光區域之反射率超過60%。因此,為了利用該特定的波長區域中光的吸收及反射來判斷植物的植被狀況及活性度,並以在-1~+1之間標準化的數值進行計算。<Activity, water content, nitrogen amount, fiber amount, and protein content of the pods> This inspection uses the vegetation index. That is, the vegetation index (especially the normalized difference vegetation index) is an indicator for using the reflected light on the surface of the plant to judge the vegetation state or activity of the plant. That is, the chlorophyll contained in the plant has a reflectance of less than 20% in a red visible region of 0.5 to 0.7 μm, and a reflectance of more than 60% in a near-infrared region of 0.7 to 1.3 μm. Therefore, in order to utilize the absorption and reflection of light in the specific wavelength region, the vegetation state and activity of the plant are judged, and the values normalized between -1 and +1 are calculated.
又,對於在熱或酸性環境下葉綠素較不穩定,因光的光分解作用所導致,尤其是因紫外線的分解能力所導致較為顯著。因此,藉由計算上述植被指數,可推測該葉緣素的含有量。藉此,可判斷活性度、收成後經過的時間(亦即為鮮度)等。Moreover, chlorophyll is less stable in a hot or acidic environment, and is caused by the photodecomposition of light, especially due to the ability of ultraviolet light to decompose. Therefore, by calculating the above vegetation index, the content of the leaf margin can be estimated. Thereby, the degree of activity, the elapsed time after the harvest (that is, the freshness) can be determined.
接著,上述植被指數與植物的氮量、纖維量、以及蛋白質含量之間具有相關性。例如紅色區域波長與近紅外光波長之2個波長區域中植物的活性度與蛋白質含量之間具有相關性。更具體而言,包含在植物裡的葉綠素含量與蛋白質含量之間具有較高的正相關性。藉此,計算植被指數並推測葉綠素含量,從而可推測蛋白質含量,並可推測並篩選作為對象的豆莢的食用品質。Next, there is a correlation between the above vegetation index and the nitrogen, fiber, and protein content of the plant. For example, there is a correlation between the activity of the plant and the protein content in the two wavelength regions of the red region wavelength and the near-infrared light wavelength. More specifically, there is a high positive correlation between the chlorophyll content contained in plants and the protein content. Thereby, the vegetation index is calculated and the chlorophyll content is estimated, so that the protein content can be estimated, and the edible quality of the target pod can be estimated and screened.
又,氮量、纖維量與前述植被指數之間具有相關性。亦即,利用紅色光區域、近紅外光區域的光學資料來計算植被指數,根據該植被指數推測氮量、纖維量,藉此可推測作為對象之豆莢的成長狀況、或營養狀態等,可藉此進行篩選。Further, there is a correlation between the amount of nitrogen and the amount of fiber and the aforementioned vegetation index. That is, the vegetation index is calculated by using the optical data of the red light region and the near-infrared light region, and the nitrogen amount and the fiber amount are estimated based on the vegetation index, thereby estimating the growth state or the nutritional state of the target pod, etc. This is filtered.
接著,上述紅色影像中紅色波長區域的反射率與近紅外光影像中近紅外光波長區域的反射率係取得由數位攝影機等影像感測器所測量出的反射亮度、反射頻譜的強度和數位值,並使用所取得的數值來計算植被指數,藉此可判斷如植被的活性、植被率、葉面積指數(LAI)之植被參數。Then, the reflectance of the red wavelength region in the red image and the reflectance of the near-infrared light wavelength region in the near-infrared light image acquire the reflected luminance, the intensity and the digital value of the reflected spectrum measured by an image sensor such as a digital camera. And use the obtained values to calculate the vegetation index, thereby judging the vegetation parameters such as vegetation activity, vegetation rate, and leaf area index (LAI).
此外,由上述紅色波長區域的反射光所構成的紅色影像、以及由近紅外光波長區域的反射光所構成的近紅外光影像係從攝影手段拍攝豆莢的反射影像抽出並取得各個波長區域的頻譜。除此之外,亦可將濾色器設置於攝影手段,並拍攝使用前述波長區域內的反射影像。藉此,例如對於一般商用數位攝影機設置可見光遮斷、紅外線光遮斷之二種種類的透鏡過濾器,亦可拍攝可見光紅色與近紅外光的影像。Further, the red image composed of the reflected light in the red wavelength region and the near-infrared light image composed of the reflected light in the near-infrared light wavelength region are extracted from the reflection image of the pod by the photographing means, and the spectrum of each wavelength region is obtained. . In addition to this, a color filter may be provided to the photographing means, and the reflected image in the wavelength region described above may be photographed. Thereby, for example, a general-purpose digital camera is provided with two types of lens filters of visible light blocking and infrared light blocking, and images of visible red and near-infrared light can also be captured.
進一步,較理想為,前述影像分析手段係對影像取得手段所取得紅色影像及近紅外光影像各別進行二值化處理,使用經二值化處理的紅色影像及近紅外光影像來計算前述植被指數。藉由二值化處理影像,再根據訂定的閥值將具有濃淡的影像轉換為黑與白二種色階,因而可容易從該影像進行檢測對象的抽出。亦即,藉由進行二值化處理,可容易進行檢測對象的抽出,因此之後計算前述植被指數等之情形時處理亦變得容易,而可高速進行處理(或執行)。Further, preferably, the image analysis means binarizes each of the red image and the near-infrared light image obtained by the image acquisition means, and uses the binarized red image and the near-infrared light image to calculate the vegetation. index. By binarizing the image, and converting the image having the shading into two levels of black and white according to the predetermined threshold, the extraction of the object to be detected can be easily performed from the image. In other words, by performing the binarization processing, the extraction of the detection target can be easily performed. Therefore, when the vegetation index or the like is calculated, the processing becomes easy, and the processing can be performed (or executed) at a high speed.
又,較理想為,前述紅色影像及近紅外光影像分別劃分成複數個區域,前述影像分析手段係計算每個被劃分區域的植被指數,對拍攝前述豆莢之整體區域計算每個區域算出的植被指數的平均值。劃分區域為在攝影手段所拍攝的影像中拍攝到莢果的區域,因此,較理想為抽出拍攝有該豆莢的區域。分折攝影影像時,較理想為,對含有複數個豆莢之攝影影像進行各莢果抽出處理,因此對取得的攝影影像進行標記處理並特定豆莢的輪廓形狀,再分別對每個豆莢創建影像。Further, preferably, the red image and the near-infrared light image are respectively divided into a plurality of regions, and the image analysis means calculates a vegetation index of each divided region, and calculates vegetation calculated for each region of the entire region of the pod. The average of the index. The divided area is an area in which the pod is photographed in the image taken by the photographing means, and therefore, it is preferable to extract the area in which the pod is photographed. In the case of splitting the photographic image, it is preferable to perform the respective pod extraction processing on the photographic image including the plurality of pods. Therefore, the obtained photographic image is subjected to labeling processing to specify the contour shape of the pod, and an image is created for each pod.
計算每個劃分區域的植被指數,藉此可對拍攝的每個豆莢進行檢查、判斷黑點、變色、或裂痕等之不良。亦即,並以在-1~+1之間標準化的數值來計算每個劃分區域,因此可辯識前述不良狀況。例如可辯識黑點的大小、變色的範圍、或是有無裂痕等。藉此計算每個劃分區域的植被指數,無需丟棄所有被辯識為不良品的豆莢,例如根據不良狀況更改出貨端等,可不浪費且利用培育的豆莢。The vegetation index of each divided area is calculated, whereby each of the photographed pods can be inspected to judge the defects of black spots, discoloration, or cracks. That is, each of the divided regions is calculated with a value normalized between -1 and +1, so that the above-described undesirable condition can be recognized. For example, the size of the black spot, the range of discoloration, or the presence or absence of cracks can be recognized. By calculating the vegetation index of each divided area, it is not necessary to discard all the pods that are identified as defective products, for example, changing the shipping end according to the bad condition, and the cultivated pods can be utilized without wasting.
接著,計算每個區域算出的植被指數的平均值,藉此可判斷作為檢查、判斷對象之豆莢的活性度、收成後經過的時間(亦即鮮度)等。亦即,取得每個劃分區域的植被指數之平均值,藉此可將1或複數個毛豆在拍攝整體區域中辯識成一個群組,並計算每個區域算出的植被指數的平均值,以便計算該一個群組的植被指數。藉由計算每個區域算出的植被指數的平均值,可作為該群組的植被指數來進行辯識,並檢查、判斷該群組內豆莢的活性度、含水量、氮量、纖維量、蛋白質含量中任一個評價項目。藉此根據鮮度、食用品質、營養狀態等篩選作為對象的豆莢。Then, the average value of the vegetation index calculated for each region is calculated, whereby the activity of the pod as the object to be inspected and judged, the elapsed time after the harvest (that is, the freshness), and the like can be determined. That is, an average of the vegetation indices of each divided area is obtained, whereby one or a plurality of edamame can be identified as a group in the overall area of the photograph, and the average value of the vegetation index calculated in each area is calculated so that Calculate the vegetation index of the one group. By calculating the average value of the vegetation index calculated in each region, it can be used as the vegetation index of the group to identify and judge the activity, water content, nitrogen amount, fiber amount and protein of the pods in the group. Any of the evaluation items. Thereby, the pods to be targeted are screened according to freshness, food quality, nutritional status, and the like.
<豆莢的裂痕檢查> 又較理想為,為了更容易檢查、判斷豆莢有無裂痕,而利用由可見光區域的反射光所構成可見光影像(以上簡稱『RGB影像』)。亦即,亦可利用前述影像取得手段來取得由可見光區域的反射光所構成的可見光影像。在前述影像分析手段中,利用從可視光區域的反射光減去紅色影像中紅色波長區域的反射率,再加上前述植被指數的數值所得之值。例如檢查、判斷毛豆的裂痕時,由綠色與白色二色構成影像,有裂痕的部位會以白色表示,故視覺上亦容易判斷。在單純顯示前述植被指數之影像中,毛豆的綠色部紅色反射量較少,非綠色部中紅色反射量較多。亦即,毛豆的裂痕部位、以及毛豆以外的部分皆以紅色顯示,因此裂痕的判斷較為困難。因此,藉由從可見光影像中可見光區域的反射率減去紅色影像中紅色波長區域的反射率,可減少紅色反射量,再加上植被指數的數值(較理想為特別經二值化處理後的數值),藉此可將非綠色部分的裂痕部分以白色顯示。因此,容易以肉眼辯識。此外,有關裂痕的判斷,不劃分如前述般的區域,可對拍攝區域整體計算植被指數,再利用上述方法以確認顏色差異之方式來檢查、判斷。<Crack Detection of Pod> It is preferable to use a visible light image (hereinafter referred to as "RGB image") which is reflected by the reflected light in the visible light region in order to make it easier to inspect and determine whether or not the pod is cracked. In other words, the visible light image formed by the reflected light in the visible light region can be obtained by the image capturing means. In the image analysis means, the reflectance of the red wavelength region in the red image is subtracted from the reflected light of the visible light region, and the value of the vegetation index is added. For example, when examining and judging the crack of edamame, the image is composed of green and white, and the cracked part is indicated by white, so it is easy to judge visually. In the image showing the vegetation index alone, the amount of red reflection in the green portion of the edamame is small, and the amount of red reflection in the non-green portion is large. That is, the cracked part of the edamame and the parts other than the edamame are displayed in red, so the judgment of the crack is difficult. Therefore, by subtracting the reflectance of the red wavelength region in the red image from the reflectance of the visible light region in the visible light image, the amount of red reflection can be reduced, and the value of the vegetation index is added (preferably after the binarization process) Numerical value, whereby the cracked portion of the non-green portion can be displayed in white. Therefore, it is easy to identify with the naked eye. Further, regarding the determination of the crack, the vegetation index is calculated for the entire photographing area without dividing the region as described above, and the above method is used to check and judge the color difference.
<豆莢篩選裝置> 接著,在本發明中,組合上述構成的豆莢篩選系統與豆莢的搬送手段等,藉此可作為連續篩選豆莢之豆莢篩選裝置。亦即,為了解決前述課題中之至少一課題,係提供一種豆莢篩選裝置,具備:搬送手段,係搬送檢查對象物;攝影手段,拍攝前述搬送手段所搬送的檢查對象物;以及豆莢篩選手段,取得前述攝影手段所拍攝的影像,同時分析前述影像。<Poisonous Seed Screening Apparatus> Next, in the present invention, the pod screening system and the pod transfer means and the like of the above-described configuration are combined, whereby the pod pod screening apparatus for continuously screening the pods can be used. In other words, in order to solve at least one of the problems, the present invention provides a pod screening device comprising: a transport means for transporting an object to be inspected; an imaging means for capturing an object to be inspected by the transport means; and a pod screening means; The image captured by the aforementioned photographing means is obtained, and the image is analyzed at the same time.
根據豆莢篩選裝置,利用搬送手段連續搬送作為檢查對象的豆莢,並以攝影手段拍攝豆莢,再利用本發明之豆莢篩選系統進行分析,藉此找出規格外的豆莢,依所需進行排除。According to the pod screening device, the pods to be inspected are continuously conveyed by the transport means, and the pods are photographed by the photographing means, and then analyzed by the pod screening system of the present invention, thereby finding the pods outside the specifications and eliminating them as needed.
又,在本發明之豆莢篩選裝置中,較理想為,進一步具備:排出手段,係以搬送手段搬送的檢查對象物內,對豆莢篩選手段中被指定的檢查對象物噴吹空氣,並吹走該檢查對象物。此時,前述豆莢篩選手段中影像分析手段係從尺寸檢查結果來特定長度與寬度的中心,該尺寸檢查係從攝影影像測量各個豆莢的長度與寬度。因此,較理想為,該排出手段係朝豆莢篩選手段中被指定為檢查對象物之長度與寬度的中心噴吹空氣。Furthermore, in the pod sorting apparatus of the present invention, it is preferable to further include: a discharge means for injecting air into the object to be inspected in the pod screening means, and blowing away the inside of the object to be inspected by the transport means The object to be inspected. At this time, the image analysis means in the pod screening means specifies the center of the length and the width from the result of the dimensional inspection, and the size inspection measures the length and width of each pod from the photographic image. Therefore, it is preferable that the discharge means blows air toward the center of the length and width of the object to be inspected in the pod screening means.
藉由設置排出手段,可將豆莢篩選系統被判斷為規格外的豆莢從搬送線上排出,可僅篩選符合規格的豆莢。又,構成該豆莢篩選系統之影像分析手段係從尺寸檢查結果來特定長度與寬度的中心,該尺寸檢查係從攝影影像測量各個豆莢的長度與寬度;藉此,前述排出手段係朝豆莢篩選手段中被指定為檢查對象物之長度與寬度的中心噴吹空氣。藉此,藉由對規格外的豆莢中心噴吹空氣,可確實排出規格外的豆莢。因此,即便在需要大量篩選之情形時,可確實且精確地檢查、篩選收成的豆莢。By providing the discharge means, the pods of the pod screening system can be judged to be discharged from the transport line, and only the pods meeting the specifications can be screened. Further, the image analysis means constituting the pod screening system is a center of a specific length and width from the result of the dimensional inspection, and the size inspection measures the length and width of each pod from the photographic image; thereby, the discharging means is directed to the pod screening means The center is designated to inject air into the center of the length and width of the inspection object. Thereby, the pods outside the specification can be reliably discharged by blowing air to the center of the pod outside the specification. Therefore, even when a large amount of screening is required, the harvested pods can be accurately and accurately inspected.
[發明功效] 依據本發明之毛豆篩選系統,可根據拍攝照射在豆莢上光的反射光之影像來檢查殼內豆子的容納狀態(尤其是豆粒數及大小),從而減去透射光的拍攝需求。又,在為了品質管理而儲存所拍攝的影像時,若能儲存拍攝反射光之影像,則可大幅減少影像的儲存區域。接著在檢查中,即便收成時間被限定固定期間而必須在短期間檢查、篩選大量豆莢時,只要將拍攝反射光之1種種類的影像進行分析,即可迅速且大量進行收成後的檢查、篩選。[Effects of the Invention] According to the edamame screening system of the present invention, the image of the beans in the shell can be inspected according to the image of the reflected light that is irradiated on the pods (especially the number and size of the beans), thereby reducing the need for transmitted light. . Moreover, when the captured image is stored for quality management, if the image of the reflected light can be stored, the image storage area can be greatly reduced. In the inspection, even if the harvest time is limited to a fixed period, it is necessary to examine and screen a large number of pods in a short period of time. As long as one type of image of the reflected light is analyzed, the inspection and screening after the harvest can be performed quickly and in large quantities. .
又,若根據攝影手段所拍攝的豆莢之反射影像、亦即由紅色波長區域的反射光所構成的紅色影像以及由近紅外光波長區域的反射光所構成的近紅外光影像來計算植被指數,則可利用該植被指數來檢查、篩選豆莢的刮痕、活性度等。亦即,可更正確地執行外觀的檢查、篩選,尤其是等同目視般檢查、篩選刮痕等。Further, the vegetation index is calculated based on the reflected image of the pod photographed by the photographing means, that is, the red image composed of the reflected light of the red wavelength region and the near-infrared light image composed of the reflected light of the near-infrared wavelength region. The vegetation index can be used to inspect and screen the scars, activity, and the like of the pods. That is, the inspection and screening of the appearance can be performed more correctly, especially the visual inspection, the screening of the scratches, and the like.
又,利用前述植被指數,不僅豆莢的外觀(尤其是裂痕),亦可對豆莢的活性度、含水量、氮量、纖維量、蛋白質含量中至少一個進行檢查或篩選。藉此,為了詳細辯識收成後經過的時間、豆莢的乾燥度(含水量)等,而更徹底執行品質管理。Further, by using the aforementioned vegetation index, not only the appearance of the pod (especially a crack) but also at least one of the activity, water content, nitrogen amount, fiber amount, and protein content of the pod can be inspected or screened. In this way, quality management is more thoroughly performed in order to recognize in detail the time elapsed after the harvest, the dryness (water content) of the pods, and the like.
接著,利用豆莢篩選裝置可迅速且大量進行收成後豆莢的檢查、篩選,該豆莢篩選裝置係使用本發明之豆莢篩選系統。亦即,在本發明中,使用反射光影像來計算植被指數,藉此對殼內中豆子的容納狀態(尤其是豆粒數或大小)、殼的缺口或有無蒂頭、豆莢的活性度、含水量、氮量、纖維量、蛋白質含量進行檢查。藉此,減去將收成後檢查、篩選劃分階段之需求。因此,利用本發明之豆莢篩選裝置,即便收成時間被限定固定期間而必須在短期間檢查、篩選大量豆莢之情形,仍可迅速且大量進行收成後的檢查、篩選。Next, the pod screening device can quickly and intensively perform inspection and screening of the pod after harvesting, and the pod screening device uses the pod screening system of the present invention. That is, in the present invention, the reflected light image is used to calculate the vegetation index, thereby accommodating the state of the beans in the shell (especially the number or size of the beans), the gap of the shell or the presence or absence of the pedicle, the activity of the pod, and The amount of water, nitrogen, fiber, and protein were examined. In this way, the need to check and screen the stages after the harvest is subtracted. Therefore, according to the pod screening device of the present invention, even when the harvest time is limited to a fixed period, it is necessary to inspect and screen a large number of pods in a short period of time, and the inspection and screening after the harvest can be performed quickly and in a large amount.
以下,參照圖示並對本發明之豆莢篩選系統之一實施形態進行具體說明。尤其是在本實施形態中,具體說明關於篩選毛豆之豆莢篩選系統。Hereinafter, an embodiment of the pod screening system of the present invention will be specifically described with reference to the drawings. In particular, in the present embodiment, a pod screening system for screening edamame is specifically described.
圖1係顯示使用本實施形態之豆莢篩選系統(以下記載為「毛豆篩選系統40」)構成之豆莢篩選裝置(以下記載為「毛豆篩選裝置」)之整體構成圖。該毛豆篩選裝置係構成為:進料部10,係收納作為篩選對象的毛豆W,同時連續供給至篩選線;第一搬送手段20,係搬送從進料部10所供給的毛豆W;第二搬送手段30,係使該第一搬送手段20所搬送的毛豆W反轉並搬送;攝影手段21、31,係設置於各個搬送手段;毛豆篩選系統40,係分析攝影手段21、31所拍攝的影像;搬送檢測手段22、32,係設置於各搬送手段並檢測搬送速度或搬送量;排出手段23、33,係設置在各搬送手段的尾端側之毛豆W掉落部。Fig. 1 is a view showing the overall configuration of a pod screening device (hereinafter referred to as "daibu screening device") which is configured using the pod screening system of the present embodiment (hereinafter referred to as "daibu screening system 40"). The edamame screening apparatus is configured such that the feeding unit 10 stores the edamame W as a screening target and continuously supplies the whipped beans W to the screening line; the first conveying means 20 conveys the edamame W supplied from the feeding unit 10; The transport means 30 reverses and transports the edamame W conveyed by the first transport means 20; the photographing means 21, 31 are provided in each transport means; and the edamame screening system 40 is formed by the analysis photographing means 21, 31. The image detecting means 22 and 32 are provided in each of the conveying means to detect the conveying speed or the conveying amount, and the discharging means 23 and 33 are the edamame W falling parts provided on the trailing end side of each conveying means.
前述進料部10具備空間部,係收納作為篩選對象之毛豆W(豆莢)。又,將收納於進料部10內部的毛豆W供給至搬送手段用之排出構造,該搬送手段係由第一搬送手段20與第二搬送手段30所構成。排出構造為設置於該進料部10之底面的皮帶輸送機等,使收納於進料部10的毛豆W朝前方移動。然而,該進料部10若能將收納於內部的毛豆W送出到執行檢查的搬送手段則亦可為其它構成。The feeding unit 10 includes a space portion, and stores the edamame W (pod) as a screening target. Moreover, the edamame W accommodated in the inside of the feeding part 10 is supplied to the discharge structure for the conveyance means, and this conveyance means is comprised by the 1st conveyance means 20 and the 2nd conveyance means 30. The discharge structure is a belt conveyor or the like provided on the bottom surface of the feeding unit 10, and the edamame W accommodated in the feeding unit 10 is moved forward. However, the feeding unit 10 may have another configuration if it can send the edamame W stored therein to the conveying means for performing the inspection.
從前述進料部10所排出的毛豆W係供給至搬送手段,該搬送手段係由第一搬送手段20以及第二搬送手段30所構成。此時,以從進料部10掉落的毛豆W互相不重疊之方式在掉落部設置振動手段(未圖示),該振動手段係一邊使從進料部排出的毛豆W振動一邊搬送。The edamame W discharged from the feeding unit 10 is supplied to the conveying means, and the conveying means is constituted by the first conveying means 20 and the second conveying means 30. At this time, a vibration means (not shown) is provided in the falling portion so that the whipped beans W dropped from the feeding portion 10 do not overlap each other, and the vibration means transports the edamame W discharged from the feeding portion while vibrating.
前述第一搬送手段20以及第二搬送手段30係可分別藉由皮帶輸送機等來構成。從進料部10供給的毛豆W一開始藉由第一搬送手段20被搬送。較理想為,在該第一搬送手段20中具有使搬送毛豆W的方向朝固定方向對齊之對準部。對準部係例如在毛豆W的搬送方向上以既定間隔配置形成長帶狀對準板或線材(未圖示)。被搬送的毛豆W係藉由碰觸該對準板,以朝向搬送方向呈縱向之方式修正方向。Each of the first conveying means 20 and the second conveying means 30 may be constituted by a belt conveyor or the like. The edamame W supplied from the feeding unit 10 is initially conveyed by the first conveying means 20. Preferably, the first conveying means 20 has an alignment portion for aligning the direction in which the edamame W is conveyed in the fixing direction. In the alignment portion, for example, a long strip-shaped alignment plate or a wire (not shown) is disposed at a predetermined interval in the conveyance direction of the edamame W. The edamame W that is conveyed is corrected in the longitudinal direction toward the conveyance direction by touching the alignment plate.
在上述第一搬送手段20的上方設置有攝影手段21,該攝影手段係拍攝由第一搬送手段20搬送而來的毛豆W。可使用攝影機或視訊攝像機作為該攝影手段21。較理想為,該攝影手段21係以靜態影像或動態影像方式拍攝被搬送的毛豆W,並以電子資料方式取得攝影影像,以便將該攝影影像直接併入毛豆篩選系統40進行分析。接著較理想為,設置可穩定照射位於該攝影手段21的攝影範圍內毛豆W之照明。藉由使照射毛豆W的亮度固定,可使攝影影像的亮度穩定。Above the first conveyance means 20, an image pickup means 21 for photographing the edamame W conveyed by the first conveyance means 20 is provided. A camera or a video camera can be used as the photographing means 21. Preferably, the photographing means 21 photographs the transferred edamame W as a still image or a moving image, and acquires the photographic image by electronic data so as to directly incorporate the photographic image into the edamame screening system 40 for analysis. Next, it is preferable to provide illumination for stably illuminating the edamame W located in the photographing range of the photographing means 21. By fixing the brightness of the irradiated peas W, the brightness of the photographic image can be stabilized.
前述攝影手段21所拍攝的影像係傳送至由電腦等構成的毛豆篩選系統40。在前述毛豆篩選系統40中,根據取得的攝影影像來檢查作為篩選對象之毛豆W,對毛豆W的大小或豆粒數、有無缺口或蒂頭、有無黑點或變色、以及有無裂痕或活性度中至少一個進行檢查,並特定規格外的毛豆W。關於該毛豆篩選系統40的構成及處理將在後面進行說明。The image captured by the photographing means 21 is transmitted to the edamame screening system 40 constituted by a computer or the like. In the edamame screening system 40, the edamame W as a screening target is examined based on the acquired photographic image, the size of the edamame W, the number of peas, the presence or absence of a gap or a pedicle, the presence or absence of black spots or discoloration, and the presence or absence of cracks or activity. At least one of the edamame W that was inspected and specified outside the specification. The configuration and processing of the edamame screening system 40 will be described later.
第一搬送手段20中設置有用以取得其搬送速度或搬送距離之搬送檢測手段22。對毛豆W的搬送資訊與攝影影像賦予關連,藉此可對在上述毛豆篩選系統40篩選的毛豆W(規格外的毛豆W)特定搬送至何處。在搬送過程中,藉由特定經毛豆篩選系統40篩選的規格外毛豆W,可將該規格外的毛豆W排出至搬送線外。可使用旋轉編碼器(rotary encoder)、線性編碼器(linear encoder)等作為搬送檢測手段22。The first transport means 20 is provided with a transport detecting means 22 for obtaining the transport speed or the transport distance. The transfer information of the edamame W is associated with the photographic image, whereby the edamame W (the edamame W outside the specification) selected by the edamame screening system 40 can be specifically transported to where. In the transfer process, the edamame W outside the specification can be discharged to the outside of the transfer line by the specific outer edamame W screened by the specific edamame screening system 40. A rotary encoder, a linear encoder, or the like can be used as the transport detecting means 22.
在本實施形態中,前述規格外的毛豆W係從第一搬送手段20的尾端掉落時被排除。亦即,在第一搬送手段20的尾端設置有使搬送的毛豆W掉落至第二搬送手段30的掉落部,從該掉落部掉落時,藉由排出手段23將經前述毛豆篩選系統40特定的毛豆W(規格外的毛豆W)從搬送線排出。排出手段23除了以對特定的毛豆W噴吹空氣,噴吹到搬送線外之方式構成之外,亦可以可取出特定的毛豆W之構成來形成。In the present embodiment, when the edamame W other than the above specifications is dropped from the trailing end of the first conveying means 20, it is excluded. In other words, at the end of the first conveying means 20, a falling portion for dropping the transferred edama W to the second conveying means 30 is provided, and when it is dropped from the falling portion, the edamame is discharged by the discharging means 23. The edamame W (the edamame W outside the specification) specified by the screening system 40 is discharged from the conveying line. The discharge means 23 may be formed by blowing air to a specific edama W and blowing it outside the transport line, or may be formed by taking out a specific whipped bean W.
如上所述,由第一搬送手段20搬送且由毛豆篩選系統40判斷為規格外的毛豆W係由排出手段23排出,剩餘毛豆W則由第二搬送手段30搬送。較理想為,由該第二搬送手段30搬送時,使毛豆W在第一搬送手段20的搬送狀態下經反轉後進行搬送。因此,較理想為,在第一搬送手段20之後、或是在第二搬送手段30的開端(第一搬送手段20的側)設置使搬送的毛豆W反轉之反轉手段。反轉手段除了例如利用讓搬送的毛豆W掉落時的落差進行反轉以外,亦可在搬送的豆莢之任一面上利用皮帶或滾輪使旋轉力發揮作用進行反轉。As described above, the edamame W which is conveyed by the first conveying means 20 and judged to be out of specification by the edamame screening system 40 is discharged by the discharging means 23, and the remaining edamame W is conveyed by the second conveying means 30. Preferably, when the second conveyance means 30 is conveyed, the edamame W is conveyed by being reversed in the conveyance state of the first conveyance means 20. Therefore, it is preferable to provide a reversing means for inverting the transferred edama W after the first conveying means 20 or at the beginning of the second conveying means 30 (on the side of the first conveying means 20). In addition, for example, the reversing means can reverse the difference when the edamame W that is transported is dropped, and the rotation force can be reversed by the belt or the roller on either side of the transferred pod.
由該第二搬送手段30搬送的毛豆W亦與前述第一搬送手段20之搬送一樣進行篩選。亦即,以攝影手段31拍攝由第二搬送手段30搬送的毛豆W,其攝影影像在毛豆篩選系統40進行分析,從而特定規格外的毛豆W。此時,毛豆的搬送速度或距離係由旋轉編碼器所構成的搬送檢測手段32來取得。接著,該規格外的毛豆W係在搬送手段的尾端由排出手段33進行噴吹空氣來排出。由於在第二搬送手段30的處理與前述第一搬送手段20的處理一樣,因此在圖示中有關與第一搬送線相同構成會在符號加上「10」的數值,並省略其詳細說明。The edamame W transported by the second transport means 30 is also screened in the same manner as the transport of the first transport means 20. In other words, the edamame W transported by the second transport means 30 is photographed by the photographing means 31, and the photographic image is analyzed by the edamame screening system 40 to specify the edamame W outside the specification. At this time, the conveying speed or distance of the edama is obtained by the conveyance detecting means 32 which consists of a rotary encoder. Next, the edamame W outside the specification is discharged by the discharge means 33 at the trailing end of the conveying means. Since the processing of the second transporting means 30 is the same as the processing of the first transporting means 20, the numerical value of "10" is added to the symbol in the same configuration as the first transporting line, and the detailed description thereof will be omitted.
其次,參照圖2至23,對檢查該毛豆篩選裝置中搬送的毛豆之毛豆篩選系統之構成、以及其處理內容具體進行說明。Next, the configuration of the edamame screening system for the edamame transported in the edamame screening device and the processing contents thereof will be specifically described with reference to Figs. 2 to 23 .
圖2係顯示本實施形態之構成豆莢篩選系統之電腦硬碟構成之例。然而,圖2的電腦500僅例示該豆莢篩選系統的代表性構成,該豆莢篩選系統只要在執行影像分析用之演算裝置、記憶體、以及程式的範圍內,亦成以專用裝置來構成。Fig. 2 is a view showing an example of a configuration of a computer hard disk constituting the pod filtering system of the present embodiment. However, the computer 500 of FIG. 2 exemplifies only a representative configuration of the pod screening system, and the pod screening system is configured by a dedicated device as long as the calculation device, the memory, and the program for image analysis are executed.
圖2所示電腦500係構成為:包含CPU501、記憶體502、語音輸出裝置503、網路介面504、顯示控制器505、顯示器506、輸入裝置介面507、鍵盤508、滑鼠509、外部儲存裝置510、外部記錄媒體驅動裝置511、以及相互連接該等構成元件之匯流排(bus)512。The computer 500 shown in FIG. 2 is configured to include a CPU 501, a memory 502, a voice output device 503, a network interface 504, a display controller 505, a display 506, an input device interface 507, a keyboard 508, a mouse 509, and an external storage device. 510. An external recording medium drive unit 511 and a bus 512 interconnecting the constituent elements.
CPU501係控制電腦500的各構成元件的動作,在作業系統的控制下對前述毛豆篩選系統的各處理的執行進行控制,並控制其動作。記憶體502一般由作為非揮發性記憶體的ROM(Read Only Memory;唯讀記憶體)、以及作為揮發性記憶體的RAM(Random Access Memory;隨機存取記憶體)所構成,相當於記憶部。在ROM中儲存電腦500起動時執行在豆莢篩選系統上執行的程式等。在RAM中由CPU501執行,分析從攝影手段取得的影像,計算毛豆的大小或豆粒數,且暫時儲存用以判斷有無缺口或蒂頭、黑點、變色、裂痕、活性度等之程式、以及該等程式在執行中使用的資料。The CPU 501 controls the operations of the respective constituent elements of the computer 500, and controls the execution of each processing of the edamame screening system under the control of the operating system, and controls the operation thereof. The memory 502 is generally composed of a ROM (Read Only Memory) as a non-volatile memory and a RAM (Random Access Memory) as a volatile memory, and corresponds to a memory unit. . The program executed on the pod screening system, etc., is executed when the storage computer 500 is started up in the ROM. Executed by the CPU 501 in the RAM, analyzes the image obtained from the photographing means, calculates the size of the edamame or the number of beans, and temporarily stores a program for determining whether there is a gap or a pedicle, a black point, a discoloration, a crack, an activity, and the like, and The data used by the program during execution.
語音輸出裝置503為揚聲器等輸出聲音的裝置,網路介面504為用以連接與各類裝置進行資訊交換之網路520用的介面。顯示控制器505為處理由CPU501發出的描繪指令用之專用控制器,將描繪資料輸出至作為顯示部的顯示器506。顯示器506係由LCD(liquid crystal display;液晶顯示器)等所構成的顯示裝置。The voice output device 503 is a device for outputting sounds such as a speaker, and the network interface 504 is an interface for connecting a network 520 for exchanging information with various devices. The display controller 505 outputs a drawing material to the display 506 as a display unit for processing a dedicated controller for drawing instructions issued by the CPU 501. The display 506 is a display device composed of an LCD (liquid crystal display) or the like.
輸入裝置介面507係接收來自鍵盤508、滑鼠509、或觸控面板等輸出入裝置所輸入的訊號,根據其訊號模式提供既定指令至CPU501。鍵盤508、滑鼠509是在進行程式的執行或設定等操作時需要。The input device interface 507 receives a signal input from an input/output device such as a keyboard 508, a mouse 509, or a touch panel, and provides a predetermined command to the CPU 501 according to the signal mode. The keyboard 508 and the mouse 509 are required for performing an operation such as execution or setting of a program.
外部儲存裝置510亦包含本說明中記憶手段之範疇。外部儲存裝置510可由例如硬式磁碟機(HDD;hard disk drive)般的記憶裝置所構成。該裝置內記錄有上述程式與資料,執行時因應所需從裝置載入至記憶體502的RAM。The external storage device 510 also includes the scope of the memory means in this description. The external storage device 510 can be constituted by a memory device such as a hard disk drive (HDD). The program and the data are recorded in the device, and are loaded from the device to the RAM of the memory 502 as needed during execution.
外部記錄媒體驅動裝置511係存取CD(Compact Disc;光碟)、MO(Magnet-Optical Disc;磁光碟)、DVD(Digital Versatile Disc;多樣化數位光碟)等可攜式的外部記錄媒體530的記錄面,並讀取記錄於該記錄面資料的裝置。The external recording medium drive device 511 accesses a record of a portable external recording medium 530 such as a CD (Compact Disc), an MO (Magnet-Optical Disc), a DVD (Digital Versatile Disc), and the like. Face and read the device recorded on the recording surface.
本實施形態之毛豆篩選系統係使用如上所述而構成的電腦來形成,並執行以下圖3至23所示之處理。The edamame screening system of the present embodiment is formed using a computer configured as described above, and the processes shown in Figs. 3 to 23 below are executed.
《基本處理》 圖3係顯示執行關於由攝影手段取得的攝影影像之基準處理內容的流程圖,圖4係顯示該基本處理內容的處理內容圖。若執行基本處理,則最先毛豆篩選系統係從攝影手段取得拍攝到複數個毛豆的攝影影像,並記錄於記憶體(S11)。接著,如圖4(B)所示,對拍攝到複數個毛豆的影像進行每個毛豆的標記處理(S12),並抽出每個毛豆的影像。再著,對每個抽出的毛豆影像進行角度修正處理(S13)。在該角度修正處理中,各毛豆影像以長尺方向(長度方向)朝橫向之方式進行影像處理。接著,如圖4(C)所示,對經角度修正的影像進行二值化處理(S15),之後對各個檢查迅速地進行所需數值、位置的特定或取得。<<Basic Processing>> FIG. 3 is a flowchart showing the execution of the reference processing content of the photographic image acquired by the photographing means, and FIG. 4 is a view showing the processing contents of the basic processing content. When the basic processing is executed, the first edamame screening system acquires a photographic image of a plurality of edamame photographed from the photographing means and records it in the memory (S11). Next, as shown in FIG. 4(B), the image processing of each edama is performed on the image in which a plurality of edamames are photographed (S12), and the image of each edamame is extracted. Then, an angle correction process is performed on each of the extracted edamame images (S13). In the angle correction processing, each edama image is subjected to image processing in a long direction (longitudinal direction) in a lateral direction. Next, as shown in FIG. 4(C), the angle-corrected image is binarized (S15), and then the desired value, position, or acquisition is quickly performed for each inspection.
《豆粒數檢查》 圖5係顯示進行豆粒數檢查的處理內容流程圖,圖6係進行該豆粒數檢查的處理內容圖。在該豆粒數檢查中,最先取得經前述基本處理的影像(二值化處理後的影像)作為檢查影像。該檢查影像由於從搬送手段搬送每個毛豆所創建而成,故該豆粒數檢查係對全部毛豆執行。Fig. 5 is a flow chart showing the processing contents of the bean number inspection, and Fig. 6 is a processing content chart for performing the bean number inspection. In the bean number inspection, the image subjected to the above-described basic processing (the image after binarization processing) is first obtained as an inspection image. Since the inspection image is created by transporting each edama from the transport means, the inspection of the number of peas is performed on all the edamame.
當取得檢查影像時,由電腦構成的毛豆篩選系統的CPU係如圖6(A)所示,判讀在檢查影像中拍攝到毛豆之區域中長度方向的畫素,並根據該畫素計算該毛豆的長度(S52)。該計算係藉由與係數相乘而算出,該係數係將攝影影片設為實測值之數值。接著,若藉由該運算取得算出的毛豆長度,則根據該長度算出豆粒數(S54)。該豆粒數的計算係如圖6所示,除了搜尋並抽出檔案41之外,即便無搜尋檔案系統亦可預先創建程式,該檔案41係訂定與長度關係之粒數,該程式係根據長度數值回傳特定粒數的數值。When the inspection image is obtained, the CPU of the edamame screening system composed of a computer is as shown in FIG. 6(A), and the pixel in the longitudinal direction of the edamame region is captured in the inspection image, and the edamame is calculated based on the pixel. Length (S52). This calculation is calculated by multiplying the coefficient by setting the photographic film to the value of the measured value. Next, when the calculated soybean bean length is obtained by the calculation, the number of soybeans is calculated based on the length (S54). The calculation of the number of beans is shown in Fig. 6. In addition to searching and extracting the file 41, even if there is no search file system, the program can be created in advance, and the file 41 is determined by the number of grains in relation to the length, and the program is based on the length. The value returns the value of a particular number of particles.
對該計算結果所得到的豆粒數來判斷是否為1顆(S56)。若為毛豆時,1顆豆粒可能存在商品價值較低的情形,因此為了提高商品價值而篩選並去排1顆豆粒的毛豆。在判斷容納於該殼內的豆子是否為1顆裡,首先在前述豆粒斷判中,先去除被判斷為1顆的毛豆。藉此,被判斷該殼內為1顆的毛豆會被判斷成規格外商品,利用排出手段特定作為排出對象。The number of beans obtained from the calculation result is judged to be one (S56). In the case of edamame, there may be a case where the value of the granules is low, so the edamame which screens and removes one peas in order to increase the value of the medicinal products. When it is judged whether or not the beans contained in the shell are one, first, in the above-mentioned bean granule judgment, the edamame judged to be one is removed first. In this way, it is determined that the edama which is one inside the shell is judged to be a product outside the specification, and is specified as a discharge target by the discharge means.
接著,前述豆粒判斷中,被判斷殼內容納豆子為2顆以上之毛豆係根據其豆粒數來劃分區域(S58)。圖6中為顯示被判斷為2顆豆子之例。因此,檢查影像係如圖6(B)所示,在長度方向上被劃分為2個區域(A1及A2)。其次,在各個區域中,將寬度值最大之地點設定為凸部。圖6(C)中,將第一區域A1中寬度最廣之地點設定為第一凸部C1,將第二區域A2中寬度最廣之地點設作為第二凸部C2。接著分別計算凸部中之數值(寬度)(S59)。Then, in the bean granule determination, it is determined that the edamame having two or more beans in the shell is divided into regions according to the number of beans (S58). Fig. 6 shows an example in which two beans are judged to be judged. Therefore, the inspection image is divided into two regions (A1 and A2) in the longitudinal direction as shown in Fig. 6(B). Next, in each area, the location where the width value is the largest is set as the convex portion. In FIG. 6(C), the most wide portion of the first region A1 is set as the first convex portion C1, and the most wide portion of the second region A2 is set as the second convex portion C2. Next, the value (width) in the convex portion is calculated separately (S59).
其次,如圖6(D)所示,分別計算凸部(C1及C2)之間SA的寬度,如圖6(E)所示,以比各凸部的寬度還小的數值且寬度為最小值之地點設定為凹部,並計算其寬度(S60)。接著,在判斷有無凹部(S61)且可算出凹部之情形時,殼內有2顆以上豆子且符合規格之毛豆,在該豆粒數檢查中會被判斷為規格內(S62)。另一方面,無法算出凹部之情形時,判斷殼內的豆子為1顆之毛豆會被判斷為規格外,並藉由排出手段特定為排出對象。亦即,毛豆篩選系統係從殼的長度、殼內的豆子數量判斷為2顆之結果(圖7(A))來執行特定前述相同凸部及凹部處理。具體而言,將該毛豆的影像劃分為二個區域(A1及A2)(圖7(B)),在各個區域中將寬度最廣的區域分別設定為凸部(C1及C2)(圖7(C))。接著,在該凸部彼此之間,比各個凸部還窄且寬度最窄的位置設定為凹部(圖7(D))。然而,在殼內的豆子為1顆之情形時,該凹部的寬度會變成與任一個凸部的寬度一樣,因此無法設定該凹部(圖7(E))。藉此,在此種情況時,判定殼內豆子為1顆,並特定為規格外之毛豆。Next, as shown in Fig. 6(D), the width of the SA between the convex portions (C1 and C2) is calculated, as shown in Fig. 6(E), and the value is smaller than the width of each convex portion and the width is the smallest. The position of the value is set as a concave portion, and the width thereof is calculated (S60). Then, when it is determined whether or not the concave portion is present (S61) and the concave portion can be calculated, the soybeans having two or more beans in the shell and conforming to the specifications are judged to be within the specifications in the bean number inspection (S62). On the other hand, when it is not possible to calculate the concave portion, it is judged that the soybeans having one bean in the shell are judged to be outside the specification, and the discharge means is specified as the discharge target. That is, the edamame screening system performs the specific convex and concave processing described above in particular from the result of the length of the shell and the number of beans in the shell (Fig. 7(A)). Specifically, the image of the edamame is divided into two regions (A1 and A2) (FIG. 7(B)), and the regions having the widest width are set as convex portions (C1 and C2) in each region (FIG. 7). (C)). Next, between the convex portions, a position narrower than each convex portion and having the narrowest width is set as a concave portion (Fig. 7(D)). However, when the number of beans in the shell is one, the width of the concave portion becomes the same as the width of any one of the convex portions, and therefore the concave portion cannot be set (Fig. 7(E)). Therefore, in this case, it is determined that the number of beans in the shell is one, and it is specified as the edamame outside the specification.
藉由如上述的處理可執行豆粒數檢查,藉此可達成處理高速化。亦即,利用殼的長度可一次排出殼內豆子為1顆,因此可減少檢查對象,從而謀求處理高速化。接著,關於長度中規格內之毛豆,藉由進一步檢查凸部及凹部,可一邊符合長度規格一邊可實際抽出殼內的豆子為1顆的毛豆,從而大幅提升篩選的速度。因此,藉由上述處理而執行豆粒數檢查,可高速且正確篩選毛豆(殼內的豆子為1顆的毛豆)。The processing of the number of beans can be performed by the above-described processing, whereby the processing speed can be attained. In other words, since the length of the shell can be used to discharge one bean in the shell at a time, the object to be inspected can be reduced, and the processing speed can be increased. Then, by further inspecting the convex portion and the concave portion of the edamame in the medium-length specification, it is possible to actually extract the edama bean having one bean in the shell while conforming to the length specification, thereby greatly increasing the speed of screening. Therefore, by performing the above-described processing and performing the bean number inspection, the edamame (the edama bean having one bean in the shell) can be screened at high speed and correctly.
《缺口檢查》 又,前述毛豆篩選系統係藉由取得經基本處理而創建的檢查影像,可進一步執行缺口檢查。在該缺口檢查中,檢查毛豆的殼內是否有缺口,在本實施形態中,從檢查影像之毛豆寬度的變化量來判斷有無缺口。"Gap Inspection" Further, the edamame screening system can further perform the gap inspection by acquiring the inspection image created by the basic processing. In the notch inspection, it is checked whether or not there is a gap in the shell of the edamame. In the present embodiment, the presence or absence of the gap is determined from the amount of change in the width of the edama having the inspection image.
圖8係顯示該缺口檢查中處理內容的流程圖,圖9係進行該缺口檢查的處理內容圖。如圖8所示,藉由該缺口檢查的執行,來讀取經基本處理而創建的檢查影像(S81)。接著,該檢查影像中寬度變化較大區域、亦即長度方向中去除兩端側的既定區域之範圍指定為檢查區域(S82)。然而,該檢查區域的指定時,除了讀取殼的寬度以外,亦可以整體長度比例進行指定,例如亦可將毛豆影像中去除整體長度約5%範圍指定為檢查區域。Fig. 8 is a flow chart showing the processing contents in the gap check, and Fig. 9 is a view showing the processing contents of the gap check. As shown in FIG. 8, the inspection image created by the basic processing is read by the execution of the gap inspection (S81). Then, the area in which the width change is large in the inspection image, that is, the range in which the predetermined areas on both end sides are removed in the longitudinal direction is designated as the inspection area (S82). However, when the inspection area is specified, in addition to the width of the reading shell, the overall length ratio may be specified. For example, a range of about 5% of the overall length of the edamame image may be designated as the inspection area.
指定檢查區域後,在該檢查範圍內指定成為檢查基準之任意位置(S83),並計算該基準位置的寬度,將其設定為基準寬度(S84)。接著,將從該基準位置在任意訂定的方向上以既定量移動之地點的寬度設定為比較寬度,依序取得該比較寬度。在本實施形態中,依序取得從基準位置朝左側逐一以既定量(例如2畫素)移動之位置的寬度設作為比較寬度(S85)。接著,計算所取得的基準寬度與各個比較寬度之差分(S86),將其差分與預先設定的容許值進行比較(S87)。計算該基準寬度與各個比較寬度之差分時,除了單純計算基準寬度與各個比較寬度之差分之外,亦可以各個比較值來計算差分變化量。在此情形時,預先設定關於差分變化量的容許值,當變化量超過容許值時進行規格外的判定。When the inspection area is designated, an arbitrary position to be the inspection standard is specified in the inspection range (S83), and the width of the reference position is calculated and set as the reference width (S84). Next, the width of the position where the quantitative movement is performed in an arbitrary direction from the reference position is set as the comparison width, and the comparison width is sequentially obtained. In the present embodiment, the width of the position shifted by the predetermined amount (for example, two pixels) from the reference position to the left side is sequentially set as the comparison width (S85). Next, the difference between the obtained reference width and each comparison width is calculated (S86), and the difference is compared with a predetermined allowable value (S87). When calculating the difference between the reference width and each of the comparison widths, in addition to simply calculating the difference between the reference width and the respective comparison widths, the difference variation amount may be calculated for each comparison value. In this case, the allowable value for the difference change amount is set in advance, and when the change amount exceeds the allowable value, the determination outside the specification is performed.
上述差分與容許值的比較結果,在該差分超過容許值時,與基準值相比表示寬度突然變窄,因此判斷該比較值處存在缺口。因此,該差分超出容許值(預設的容許值)時,進行規格外的判定(S89)。另一方面,在根據該檢查區域中任意的基準位置所算出的比較寬度之差分為容許值以下時,在該檢查區域設定下一個基準位置(S90)。亦即,將仍未檢查的範圍設定為新基準位置。新基準位置為在先前設定的基準位置上僅加上任意數值且朝長度方向移動之位置,將該位置作為新基準位置,再計算與比較寬度之差分,並將其差分值與容許值進行比對,判斷規格內或規格外。豆莢內,尤其是毛豆,會因殼內是否存在豆子而其寬度會大幅變化。因此,在任一個地點作為基準位置中,比較值之差分亦會大幅變化。其結果,判斷前述凸部與凹部是否為正常形狀的變化、或是是否有缺口皆為困難。因此,本實施形態中,檢查1個毛豆時,構成為:在檢查有無缺口時所設定的檢查區域內設定複數個基準位置,並分別計算每個基準位置的寬度與比較寬度間之差分,並檢查是否存在缺口。As a result of the comparison between the difference and the allowable value, when the difference exceeds the allowable value, the width is suddenly narrowed compared to the reference value. Therefore, it is determined that there is a gap in the comparison value. Therefore, when the difference exceeds the allowable value (predetermined allowable value), the determination outside the specification is performed (S89). On the other hand, when the difference between the comparison widths calculated based on the arbitrary reference position in the inspection region is equal to or less than the allowable value, the next reference position is set in the inspection region (S90). That is, the range that has not been checked is set as the new reference position. The new reference position is a position where only an arbitrary value is added to the previously set reference position and moved in the longitudinal direction, and the position is taken as a new reference position, and the difference from the comparison width is calculated, and the difference value is compared with the allowable value. Yes, judge the specification or outside the specification. Within the pods, especially edamame, the width of the beans will vary greatly due to the presence of beans in the shell. Therefore, in any of the locations as the reference position, the difference in the comparison value also changes greatly. As a result, it is difficult to determine whether or not the convex portion and the concave portion are changed in the normal shape or whether there is a gap. Therefore, in the present embodiment, when one edama is inspected, a plurality of reference positions are set in the inspection area set when the presence or absence of the inspection is detected, and the difference between the width of each reference position and the comparison width is calculated. Check for gaps.
接著,如上所述一邊朝長度方向移動位置一邊對每個設定的基準位置進行比較寬度的差分與容許值的比對。接著,針對所有檢查區域,在對比結束之時序上,在所有基準位置上,若比較寬度之差分為容許值以下則判斷為規格內,並結束執行毛豆的缺口檢查。Next, as described above, the position of the relative width is compared with the allowable value for each set reference position while moving the position in the longitudinal direction. Then, for all the inspection areas, at the timing of the comparison, if the difference of the comparison width is equal to or less than the allowable value at all the reference positions, it is judged to be within the specification, and the inspection of the gap of the edama is completed.
根據以上的處理內容判斷毛豆中有無缺口,藉此即便為寬度變化較大的毛豆,仍可正確地檢測缺口。又,因訂定基準位置,且僅計算比較寬度之差分,因此其處理速度可高速化,亦可使每個單位時間的毛豆篩選量變多。Based on the above processing contents, it is judged whether or not there is a gap in the edamame, whereby the nick is accurately detected even in the case of edama having a large change in width. Further, since the difference between the comparison widths is calculated by setting the reference position, the processing speed can be increased, and the amount of edamame screening per unit time can be increased.
《蒂頭檢查》 又,前述毛豆篩選系統係取得經基本處理而創建的檢查影像,可進一步執行蒂頭檢查。該蒂頭檢查中,檢查從毛豆的殼上是否有突出的蒂頭。在本實施形態中,利用檢查影像中毛豆寬度方向的邊緣數量來判斷有無蒂頭。此外,該邊緣係指寬度方向中輪廓的邊界點,在檢查影像中可特定毛豆影像與背景顏色的色值變化的地點。"Titus Inspection" Further, the aforementioned edamame screening system obtains an inspection image created by basic processing, and can further perform a pedicle inspection. In the pedicel inspection, check if there is a protruding pedicle on the shell of the edamame. In the present embodiment, the presence or absence of the pedicle is determined by checking the number of edges in the width direction of the edama in the image. Further, the edge refers to a boundary point of the contour in the width direction, and a place where the color value of the edamame image and the background color change can be specified in the inspection image.
圖10係顯示該蒂頭檢查中處理內容的流程圖,圖11係進行該蒂頭檢查的處理內容圖。如圖10所示,利用該蒂頭檢查的執行指令,來讀取在基本處理創建的檢查影像(S101)。接著關於所取得的檢查影像如圖11(B)所示,檢測長度方向上邊緣的數量(S102)。具體而言,在檢查影像中毛豆的影像裡,檢測寬度方向的輪廓上存在有點(以下記載為「邊緣」)的數量。接著,判斷所檢測的邊緣數量(S103)。判斷該邊緣數量時如圖11(C)的符號E2所示,若邊緣為2個,則表示毛豆的正常殼的輪廓。因此,未存在該蒂頭之毛豆將判斷為規格內(S105)。另一方面,如符號E1所示,邊緣數量為3個以上,具體而言為4個時,除了殼的輪廓以外仍存在突出部分。由於該突出部分相當於毛豆的蒂頭,因此當邊緣數量為4個之地點時特定作為有蒂頭之地點,並判斷有無該蒂頭(S104)。Fig. 10 is a flow chart showing the processing contents in the head inspection, and Fig. 11 is a view showing the processing contents of the head inspection. As shown in FIG. 10, the inspection image created in the basic processing is read by the execution command of the stylus inspection (S101). Next, as shown in FIG. 11(B), the acquired inspection image is detected by the number of upper edges in the longitudinal direction (S102). Specifically, in the image of the edamame in the inspection image, the number of points in the width direction (hereinafter referred to as "edge") is detected. Next, the number of detected edges is judged (S103). When the number of edges is judged, as shown by the symbol E2 in Fig. 11(C), if the number of edges is two, the outline of the normal shell of the edamame is indicated. Therefore, the edamame in which the pedicle is not present is judged to be within the specification (S105). On the other hand, as indicated by the symbol E1, when the number of edges is three or more, specifically, four, there are still protruding portions other than the outline of the shell. Since the protruding portion corresponds to the pedicle of the edamame, when the number of the edges is four, it is specified as the location of the pedicle, and it is judged whether or not the pedicle is present (S104).
接著,關於經判斷有蒂頭之毛豆,接下來是計算該蒂頭的長度(S106)。該蒂頭長度的計算係如圖11(C)所示,計算從算出邊緣的數量為4個之地點至邊緣的數量集中到2個或3個之地點為止長度。接著將算出蒂頭的長度與預先設定的容許值進行比較(S108),蒂頭的長度為容許值以下時,判斷為規格內。另一方面,蒂頭的長度超過容許值時,判斷為規格外,並記錄作為前述排出手段的排出對象。Next, regarding the edamame having been judged to have a pedicle, the length of the pedicle is calculated next (S106). The calculation of the length of the pedicle head is as shown in Fig. 11(C), and the length from the point where the number of calculated edges is four to the point where the number of edges is concentrated to two or three places is calculated. Next, the length of the calculated tip is compared with a predetermined allowable value (S108), and when the length of the tip is equal to or less than the allowable value, it is determined to be within the specification. On the other hand, when the length of the pedicle exceeds the allowable value, it is judged that it is outside the specification, and the discharge target as the discharge means is recorded.
根據上述處理,不僅單純檢查有無蒂頭,亦可檢查該蒂頭的長度,因此可更正確地執實篩選。此外,本發明之豆莢篩選系統及豆莢篩選裝置並未限定於所示實施形態之態樣,亦可進行各種變化。例如構成為:可選擇執行豆粒數檢查、缺口檢杳、以及蒂頭檢查中之任一個,亦或構成為:將執行各處理的程式以其它邏輯構成。According to the above processing, not only the presence or absence of the pedicle but also the length of the pedicle can be checked, so that the screening can be performed more correctly. Further, the pod screening system and the pod screening device of the present invention are not limited to the embodiment shown, and various modifications are possible. For example, the configuration may be such that one of the bean number inspection, the notch inspection, and the pedicle inspection may be performed, or the program for executing each processing may be configured by another logic.
《植被指數計算處理》 圖12係顯示執行從攝影手段取得的攝影影像之植被指數計算處理內容的流程圖,圖13係顯示該植被指數計算處理內容的處理內容圖。當執行植被指數計算處理時,首先毛豆篩選系統係從攝影手段取得拍攝有複數個毛豆的攝影影像,具體而言為從攝影手段取得由紅色波長區域的反射光所構成的紅色影像(圖13中顯示為『R影像』)、與由近紅外光波長區域的反射光所構成的近紅外光影像(圖13中顯示為『IR影像』),並記錄於記憶體(S111、S121)。接著,如圖13(B)所示,對於拍攝有複數個毛豆的影像,將每一個毛豆進行標記處理(S112、S122),並抽出每一個毛豆的影像,再對每一個抽出的毛豆像進行角度修正處理(S113、S123)。在該角度修正處理中,將各毛豆的影像的長帶狀方向(長度方向)朝縱向之方式進行影像處埋。接著,如圖13(C)所示,對經角度修正的影像進行二植化處理(S114、S124),之後在各檢查中迅速進行所需數值、位置的特定或取得。亦即,使用上述經二值化處理的紅色影像及近紅外光影像,並如圖13(D)所示計算植被指數(S115),輸出作為檢查影像(S116)。此外,在本實施形態中,雖執行上述標記處理、角度修正處理、二值化處理,但即便不執行上述處理亦能計算植被指數。<< Vegetation Index Calculation Processing>> FIG. 12 is a flowchart showing the content of the vegetation index calculation processing of the photographed image acquired from the photographing means, and FIG. 13 is a view showing the processing content of the vegetation index calculation processing content. When performing the vegetation index calculation process, first, the edamame screening system obtains a photographic image in which a plurality of edamame images are taken from a photographing means, specifically, a red image composed of reflected light of a red wavelength region is obtained from a photographing means (in FIG. 13 A near-infrared light image (shown as "IR image" in FIG. 13) which is displayed as "R image" and reflected light in the near-infrared light wavelength region is recorded in the memory (S111, S121). Next, as shown in FIG. 13(B), for the image in which a plurality of edamame images are taken, each edama is subjected to labeling processing (S112, S122), and an image of each edamame is extracted, and each edamame image is extracted. Angle correction processing (S113, S123). In the angle correction processing, the image is buried in the longitudinal direction (longitudinal direction) of the image of each soybean in the longitudinal direction. Next, as shown in FIG. 13(C), the angle-corrected image is subjected to the second planting process (S114, S124), and then the desired value or position is specified or acquired quickly in each test. That is, the binarized red image and the near-infrared light image are used, and the vegetation index (S115) is calculated as shown in FIG. 13(D), and output as an inspection image (S116). Further, in the present embodiment, the above-described marking processing, angle correction processing, and binarization processing are performed, but the vegetation index can be calculated without performing the above processing.
《黑點、變色、裂痕檢查》 圖14係顯示進行黑點、變色、裂痕檢查之處理內容的流程圖,圖15係顯示黑點、變色、裂痕檢查中之指定豆殼區域/黑點區域態樣的前視圖。在黑點、變色、裂痕檢查中,首先利用攝影機等的攝影手段來取得作為對象的毛豆的檢查影像(S151),並指定取得影像中殼區域及不良區域(S152)。此外,圖15中係顯示指定毛豆的殼區域61與黑點區域62之態樣,再顯示執行黑點檢查時之處理內容。"Black Point, Discoloration, and Crack Inspection" Fig. 14 is a flow chart showing the processing contents of black spot, discoloration, and crack inspection, and Fig. 15 shows the specified bean shell area/black dot area state in black spot, discoloration, and crack inspection. Kind of front view. In the black spot, the discoloration, and the crack inspection, the inspection image of the target soybean is obtained by the photographing means such as a camera (S151), and the image in the shell region and the defective region are designated (S152). Further, in Fig. 15, the state of the shell region 61 and the black dot region 62 of the designated edama are displayed, and the processing contents at the time of performing the black dot inspection are displayed.
如上所述,指定殼區域及不良區域後,計算每個區域的植被指數(S153)。此外,在執行黑點、變色、裂痕檢查時,即便不計算植被指數亦可檢測黑點、變色、裂痕,但利用植被指數可幫助肉眼輕易辯識,之後的處理亦容易進行。As described above, after specifying the shell region and the defective region, the vegetation index of each region is calculated (S153). In addition, when performing black spot, discoloration, and crack inspection, black spots, discoloration, and cracks can be detected even without calculating the vegetation index. However, the vegetation index can be easily recognized by the naked eye, and subsequent processing is easy.
接著,計算占據殼整體的不良區域的比率(S154),並判斷不良區域的比率(S155),藉此判斷作為對象的毛豆是否為規格內。亦即,根據上述計算不良區域的比率,來判斷不良區域的比率是否在設定的容許值範圍內(S156)。作為對象的毛豆不良區域的比率為容許值內時,將作為規格內商品並搬送至下一製程(S158)。另一方面,在作為對象的毛豆不良區域的比率超過容許值時,將判斷為規格外商品並利用排出手段特定作為排出對象(S157)。Next, the ratio of the defective area occupying the entire casing is calculated (S154), and the ratio of the defective area is determined (S155), thereby judging whether or not the target soybean is within the specification. That is, it is determined whether or not the ratio of the defective area is within the set allowable value range based on the ratio of the calculated defective area (S156). When the ratio of the edama-poor area to be the target is within the allowable value, the product in the specification is transported to the next process (S158). On the other hand, when the ratio of the target edama-defective area exceeds the allowable value, it is determined as the product outside the specification and is specified as the discharge target by the discharge means (S157).
《活性度檢查》 其次,當作為對象的毛豆執行活性度檢查時,如圖16所示根據流程圖進行說明。在活性度檢查中,首先利用攝影機等攝影手段來取得作為對象的毛豆之檢查影像(S171),再指定所取得影像中的檢查區域(S172)。接著,計算每個指定區域的植被指數(S173),並計算每個區域的植被指數之平均值(S174)。在此,藉由取得每個區域的植被指數之平均值,可將1或複數個毛豆在拍攝整體區域中辯識成一個群組。藉由計算每個區域植被指數的平均值,而可辯識該群組的植被指數,亦可辯識該群組內毛豆的活性度。藉此,由於可辯識作為對象之毛豆的活性度,故可依活性度進行篩選。<<Activity Check>> Next, when the edama having the target is subjected to the activity check, it will be described based on a flowchart as shown in FIG. 16 . In the activity check, first, an inspection image of the target soybean is obtained by an imaging means such as a camera (S171), and an inspection region in the acquired image is designated (S172). Next, the vegetation index for each designated area is calculated (S173), and the average of the vegetation indices of each area is calculated (S174). Here, by taking the average of the vegetation indices of each region, one or a plurality of edamame can be identified as a group in the overall region of the photographing. By calculating the average of the vegetation indices in each region, the vegetation index of the group can be identified, and the activity of the soybeans in the group can also be identified. Thereby, since the activity of the target soybean is identifiable, it is possible to perform screening according to the activity.
接著,根據上述內容判斷算出的活性度、亦即每個區域的植被指數的平均值是否在預先設定容許值的範圍內(S175)。其結果,在作為對象之毛豆的活性度為容許值內時,將作為規格內商品並搬送至下一製程(S177)。另一方面,作為對象之毛豆的活性度超出容許值時,將判斷為規格外商品,並利用排出手段特定作為排出對象(S176)。Next, based on the above, it is judged whether or not the calculated activity degree, that is, the average value of the vegetation index of each region is within the range of the preset allowable value (S175). As a result, when the activity level of the target soybean is within the allowable value, the product in the specification is transported to the next process (S177). On the other hand, when the activity degree of the target soybean is out of the allowable value, it is determined as the product outside the specification, and is specified as the discharge target by the discharge means (S176).
《利用可見光影像之裂痕檢查》 其次,關於第2實施形態之裂痕檢查的處理內容,如圖17所示根據流程圖進行說明。在第2實施形態之裂痕檢查中,利用由可見光區域的反射光所構成的可見光影像來執行裂痕檢查。首先,取得從攝影手段拍攝複數個毛豆的攝影影像,具體而言為取得可見光影像(以下稱為「RGB影像」)以及近紅外光影像(以下稱為「IR影像」)(S181、S191),該可見光影像係拍攝可見光波長區域的反射光,該近紅外光影像係拍攝近紅外光波長區域的反射光。此外,取得上述影像時,除了從攝影手段拍攝的毛豆的反射影像抽出並取得各個波長區域的頻譜之外,亦可在攝影手段設置濾色器(color filter),拍攝且使用前述波長區域內的反射影像。<<Rift Inspection Using Visible Light Image>> Next, the processing content of the crack inspection in the second embodiment will be described with reference to a flowchart as shown in Fig. 17 . In the crack inspection of the second embodiment, the crack inspection is performed using the visible light image composed of the reflected light in the visible light region. First, a photographic image in which a plurality of edamame images are captured by a photographing means, specifically, a visible light image (hereinafter referred to as "RGB image") and a near-infrared light image (hereinafter referred to as "IR image") are obtained (S181, S191). The visible light image captures reflected light in a visible light wavelength region, and the near-infrared light image captures reflected light in a near-infrared light wavelength region. Further, when the image is obtained, in addition to extracting the reflected image of the edamame photographed by the photographing means and acquiring the spectrum of each wavelength region, a color filter may be provided in the photographing means to capture and use the wavelength region. Reflect the image.
接著,在各個影像中指定作為對象的檢查區域(S182、S192)。較理想為,指定檢查區域後,為了修正因攝影機視差導致畫素的偏差,對取得的RGB影像及IR影像執行影像合成處理(或畫素匹配(pixel matching))(S183)。Next, an inspection area as a target is specified in each image (S182, S192). Preferably, after the inspection area is designated, image composition processing (or pixel matching) is performed on the acquired RGB image and IR image in order to correct the pixel deviation caused by the camera parallax (S183).
根據上述內容,在合成影像中計算植被指數(S184),並進行二值化處理(S185)。另一方面,在合成影像中可同時進行從可見光區域的反射率減去紅色區域的反射率之製程(S193)。接著,在從可見光區域的反射率減去紅色區域的反射率所得數值上加上將植被指數進行二值化後之數值(S186),藉此可抽出以肉眼就能辯識的裂痕。According to the above, the vegetation index is calculated in the synthesized image (S184), and binarization processing is performed (S185). On the other hand, in the synthesized image, the process of subtracting the reflectance of the red region from the reflectance in the visible light region can be simultaneously performed (S193). Next, a numerical value obtained by binarizing the vegetation index is added to the value obtained by subtracting the reflectance of the red region from the reflectance in the visible light region (S186), whereby the crack which can be recognized by the naked eye can be extracted.
接著,根據上述內容來判斷被抽出的毛豆有無裂痕(S187)。其結果,在作為對象的毛豆上無裂痕時,將作為規格內商品搬送至下一製程(S189)。另一方面,在作為對象的毛豆有裂痕時,將判斷作為規格外商品,並利用排出手段特定作為排出對象(S188)。 [實施例1]Next, based on the above, it is judged whether or not the extracted edama has cracks (S187). As a result, when there is no crack on the target soybean, the product in the specification is transported to the next process (S189). On the other hand, when there is a crack in the edama as a target, it is judged as a product other than the specification, and it is specified as a discharge target by the discharge means (S188). [Example 1]
在實施例1中,驗證上述毛豆的裂痕檢查可否為肉眼能辯識的有效手法。 <使用機械> 本實施例中,為了取得毛豆的殼的4波帶影像(band image),在約1,470萬畫素的數位攝影機(CANON PowerShot G10)上去除紅外線截止濾光器(cut‐off filter)(以下稱為「IR截止濾光器」),並安裝紅外線透射濾光器(以下稱為「IR濾光器」)並取得紅外線影像(以下簡稱「IR」影像)。此外,所使用的IR濾光器為FUJIFILM製造的IR-76,78,80,82,84,86,88,90,92,94,各透射率顯示於圖18的圖表。此外,該IR濾光器為可吸收700nm以下短波長側的可見光,並穿過紅外線的濾光器。In Example 1, it was verified whether the crack test of the above-mentioned edamame can be an effective method for the naked eye to recognize. <Usage Machinery> In the present embodiment, in order to obtain a band image of a edamame shell, an infrared cut filter (cut-off filter) is removed on a digital camera (CANON PowerShot G10) of about 14.7 million pixels. (hereinafter referred to as "IR cut filter"), and an infrared transmission filter (hereinafter referred to as "IR filter") is mounted to obtain an infrared image (hereinafter referred to as "IR" image). Further, the IR filter used was IR-76, 78, 80, 82, 84, 86, 88, 90, 92, 94 manufactured by FUJIFILM, and the respective transmittances are shown in the graph of Fig. 18. Further, the IR filter is a filter that absorbs visible light on the short-wavelength side of 700 nm or less and passes through the infrared ray.
其次,在無改造數位攝影機(CANON PowerShot G10)上取得RGB的3波帶影像(以下簡稱「RGB」影像)。此二種影像以不偏移光軸之方式從試樣高度為30cm處固定的條件進行拍攝。Next, an RGB 3-band image (hereinafter referred to as "RGB" image) was obtained on a CANON PowerShot G10. These two kinds of images were taken under the condition that the height of the sample was fixed at a height of 30 cm without shifting the optical axis.
此外,在拍攝IR影像與RGB影像的4波帶影像處理中,兩個攝影機的透鏡光軸以三維空間精確地呈平行是相當困難,因此影像上以不同形式成像,亦發生因攝影機視差導致畫素偏移。因此,在執行處理上不可欠缺影像的變形修正、以及RGB影像與IR影像的合成處理(畫素匹配)。因此,進行取得影像的畫素匹配,並進行NDVI(植被指數)的計算與可視化。該NDVI(植被指數)的計算與可視化之處理係如下步驟進行。亦即,首先1)讀取處理、合成的RGB、IR影像,2)從RGB、IR影像指定測定劃分。其次,3)合成影像,4)修正經合成後影像之失真。接著,5)計算植被指數並進行可視化、圖表化。In addition, in the 4-band image processing of IR images and RGB images, it is quite difficult for the optical axes of the two cameras to be parallel in three-dimensional space. Therefore, the images are imaged in different forms, and the image is caused by camera parallax. Prime offset. Therefore, it is necessary to perform distortion correction of the image and synthesis processing (pixel matching) of the RGB image and the IR image in performing the processing. Therefore, the pixel matching of the acquired image is performed, and the calculation and visualization of the NDVI (vegetation index) are performed. The calculation and visualization of the NDVI (vegetation index) is performed as follows. That is, first, 1) read processing, synthesized RGB, and IR images, and 2) specify measurement division from RGB and IR images. Secondly, 3) synthesize the image, 4) correct the distortion of the synthesized image. Next, 5) calculate the vegetation index and visualize and graph it.
<驗證方法> 作為試驗試樣,使用毛豆(品種為「秘傳」),並備有良品、以肉眼可辯識裂痕、以及肉眼難以辯識裂痕之三種種類的豆莢。在本試驗中,首先1)將上述三種種類的試驗試樣置放在劃有四角皆為10cm之正方形的白色繪圖紙的框線內(參照圖10),其次2)一邊設置燈光並確保充分的亮度一邊以三腳架固定安裝有IR濾光器的數位攝影機,並進行拍攝。此外,從豆莢殼到數位攝影機的高度、快門速度、光圈(F值)皆以相同條件進行。接著,3)上述攝影分別以IR濾光器76~94進行,4)對取得的影像進行NDVI(植被指數)的計算以及可視化處理,再驗證可否判定毛豆的裂痕。<Verification method> As a test sample, edamame (variety is "secret") is used, and there are three types of pods which are good in quality, can be recognized by the naked eye, and are difficult to distinguish cracks by the naked eye. In this test, first, 1) place the above three types of test specimens in the frame of a white drawing paper with a square of 10 cm in all corners (see Fig. 10), and secondly, 2) set the lights and ensure sufficient The brightness is fixed on a tripod with a digital camera mounted with an IR filter and shooting. In addition, the height, shutter speed, and aperture (F value) from the pod shell to the digital camera are all performed under the same conditions. Next, 3) the above-described photographing is performed by the IR filters 76 to 94, and 4) the NDVI (vegetation index) calculation and visualization processing are performed on the acquired image, and it is verified whether the crack of the edamame can be determined.
<結果與考察> 圖20為本實驗中經拍攝取得的毛豆殼的IR影像及RGB影像,圖21為經影像處理軟體取得的NDVI(植被指數)影像。又,圖22係以NDVI(植被指數)作為基準之二值化處理影像,圖23係顯示將經二值化處理的NDVI影像覆蓋到從RGB影像減去紅色數值之影像上。<Results and Investigations> Fig. 20 is an IR image and an RGB image of the soybean shell obtained in the experiment, and Fig. 21 is an NDVI (vegetation index) image obtained by the image processing software. Further, Fig. 22 is a binarized image with NDVI ( Vegetation Index) as a reference, and Fig. 23 is a view showing that the binarized NDVI image is overlaid onto the image from which the red value is subtracted from the RGB image.
如圖21所示,在拍攝的影像上進行影像處理之NDVI影像中,以肉眼可辯識之裂痕皆能在所有影像上進行辯識。As shown in Fig. 21, in the NDVI image for image processing on the captured image, the cracks that are identifiable by the naked eye can be recognized on all images.
又, 探討針對NDVI影像執行二值化處理後是否仍有無法以肉眼辯識裂痕。其結果如圖22所示,在二值化處理中,比NDVI影像還更容易以肉眼辯識裂痕。此外,為了減輕因燈光亮度導致的影響,較理想為使用IR-86以下的濾光器,藉由使用IR-86以下的濾光器,確定可更容易辯識殼的裂痕。In addition, it is still impossible to visually recognize cracks after performing binarization processing on NDVI images. As a result, as shown in FIG. 22, in the binarization process, cracks are more easily recognized by the naked eye than NDVI images. In addition, in order to reduce the influence caused by the brightness of the light, it is preferable to use a filter of IR-86 or lower, and it is determined that the crack of the shell can be more easily recognized by using a filter of IR-86 or lower.
又,在NDVI影像中具有綠色部的紅色反射量較少,非綠色部的紅色反射量較多之性質。利用該性質,將經二值化處理的NDVI影像覆蓋到從RGB影像大幅減去紅色數值之影像上,藉此如圖23所示探討是否能辯識非綠色裂痕。其結果,以肉眼可辯識裂痕不需發生黑點,就比經二值化處理的NDVI影像還更容易進行確認。Further, in the NDVI image, the amount of red reflection in the green portion is small, and the amount of red reflection in the non-green portion is large. Using this property, the binarized NDVI image is overlaid onto the image from which the red value is greatly subtracted from the RGB image, thereby as to whether or not the non-green crack can be identified as shown in FIG. As a result, the black spots can be recognized by the naked eye, and it is easier to confirm than the binarized NDVI image.
根據上述內容,在上述驗證中將經二值化處理的NDVI影像覆蓋到從RGB影像大幅減去紅色數值之處理為比單以NDVI影像、以及將其二值化處理還容易以肉眼辯識確認裂痕。亦即,藉由上述處理在篩選毛豆上,可正確且迅速進行檢查、篩選。According to the above, in the above verification, the binarized NDVI image is overwritten to substantially reduce the red value from the RGB image, and the NDVI image is more than the NDVI image alone, and the binarization process is easily recognized by the naked eye. crack. That is to say, by screening the edama beans by the above treatment, the inspection and screening can be performed correctly and promptly.
10‧‧‧進料部
20‧‧‧第一搬送手段
21、31‧‧‧攝影手段
22、32‧‧‧搬送檢測手段
23、33‧‧‧排出手段
30‧‧‧第二搬送手段
40‧‧‧毛豆篩選系統
W‧‧‧毛豆10‧‧‧Feeding Department
20‧‧‧First means of transport
21, 31‧‧‧ means of photography
22, 32‧‧‧Transmission detection means
23, 33‧‧‧ means of discharge
30‧‧‧Second transport means
40‧‧‧Maodou screening system
W‧‧‧ edamame
圖1係顯示使用本實施形態之毛豆篩選系統構成之毛豆篩選裝置的整體構成圖。 圖2係顯示硬體構成的方塊圖。 圖3係顯示基本處理內容的流程圖。 圖4係顯示基本處理內容的處理內容圖。 圖5係顯示進行豆粒數檢查之處理內容的流程圖。 圖6係進行豆粒數檢查的處理內容圖。 圖7係顯示1顆豆子檢查之處理內容的流程圖。 圖8係顯示缺口檢查中之處理內容的流程圖。 圖9係進行缺口檢查的處理內容圖。 圖10係顯示蒂頭檢查中之處理內容的流程圖。 圖11係進行蒂頭檢查的處理內容圖。 圖12係顯示植被指數計算處理內容的流程圖。 圖13係顯示植被指數計算處理內容的處理內容圖。 圖14係顯示進行黑點、變色、裂痕檢查之處理內容的流程圖。 圖15係顯示黑點、變色、裂痕檢查中之指定豆殼區域/黑點區域態樣的前視圖。 圖16係顯示活性度檢查之處理內容的流程圖。 圖17係顯示第2實施形態之裂痕檢查處理內容的流程圖。 圖18係顯示在實施例1使用之紅外線濾光片的透射率之圖表。 圖19係顯示在實施例1使用3種試驗試樣的立體圖。 圖20係實施例1中由攝影取得毛豆莢的IR影像及RGB影像。 圖21係實施例1中由影像處理軟體取得的NDVI(植被指數)影像。 圖22係將實施例1中NDVI(植被指數)作為基準之二值化(binarization)處理影像。 圖23係將經二值化處理的NDVI影像覆蓋到實施例1中從RGB影像減去紅色數值之影像上。Fig. 1 is a view showing the overall configuration of a edamame screening apparatus constructed using the edamame screening system of the present embodiment. Figure 2 is a block diagram showing the construction of a hardware. Figure 3 is a flow chart showing the basic processing contents. Fig. 4 is a view showing the processing contents of the basic processing contents. Fig. 5 is a flow chart showing the processing contents of the bean number inspection. Fig. 6 is a view showing the processing contents of the bean number inspection. Fig. 7 is a flow chart showing the processing contents of one bean inspection. Fig. 8 is a flow chart showing the processing contents in the gap check. Fig. 9 is a view showing the processing contents of the notch inspection. Fig. 10 is a flow chart showing the processing contents in the pedicle inspection. Fig. 11 is a view showing the processing contents of the pedicle inspection. Fig. 12 is a flow chart showing the contents of the vegetation index calculation processing. Fig. 13 is a view showing the processing contents of the vegetation index calculation processing contents. Fig. 14 is a flow chart showing the processing contents of black spot, discoloration, and crack inspection. Figure 15 is a front elevational view showing the specified bean shell area/black dot area pattern in black spot, discoloration, and crack inspection. Fig. 16 is a flow chart showing the processing contents of the activity check. Fig. 17 is a flow chart showing the contents of the crack inspection processing in the second embodiment. Fig. 18 is a graph showing the transmittance of the infrared filter used in Example 1. Fig. 19 is a perspective view showing the use of three test specimens in Example 1. Fig. 20 is an IR image and an RGB image obtained by photographing the pea pods in the first embodiment. Figure 21 is an NDVI (vegetation index) image obtained by the image processing software in the first embodiment. Fig. 22 is a binarization processed image in which the NDVI (vegetation index) in Example 1 is used as a reference. Fig. 23 is a view in which the binarized NDVI image is overlaid on the image in which the red value is subtracted from the RGB image in the first embodiment.
Claims (12)
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2015031088 | 2015-02-19 | ||
JP2015-031088 | 2015-02-19 | ||
JP2015066102 | 2015-03-27 | ||
JP2015-066102 | 2015-03-27 |
Publications (2)
Publication Number | Publication Date |
---|---|
TW201636596A TW201636596A (en) | 2016-10-16 |
TWI632360B true TWI632360B (en) | 2018-08-11 |
Family
ID=56692534
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW105104823A TWI632360B (en) | 2015-02-19 | 2016-02-18 | Pod screening system and pod screening device |
Country Status (3)
Country | Link |
---|---|
JP (1) | JP6590907B2 (en) |
TW (1) | TWI632360B (en) |
WO (1) | WO2016133175A1 (en) |
Families Citing this family (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018175555A1 (en) * | 2017-03-21 | 2018-09-27 | Monsanto Technology Llc | Seed sorting |
CN107350176A (en) * | 2017-07-18 | 2017-11-17 | 商洛学院 | A kind of corn seed quality detection device |
JP6621215B2 (en) * | 2017-12-19 | 2019-12-18 | 株式会社茶野電機工業 | Green soybean sorting machine |
TWI646325B (en) * | 2018-03-12 | 2019-01-01 | 朝陽科技大學 | Pill automatic optical detection device and detection method |
US11673166B2 (en) * | 2018-03-14 | 2023-06-13 | Monsanto Technology Llc | Seed imaging |
MX2020013129A (en) | 2018-06-11 | 2021-02-18 | Monsanto Technology Llc | Seed sorting. |
TWI662481B (en) * | 2018-06-15 | 2019-06-11 | China University Of Science And Technology | Automatic measurement system for optical image processing of rice water content |
JP2020151613A (en) * | 2019-03-18 | 2020-09-24 | ヤンマーパワーテクノロジー株式会社 | Crop sorting apparatus |
JP2021043149A (en) * | 2019-09-13 | 2021-03-18 | 日世株式会社 | Detection system and detection method |
JP7354869B2 (en) * | 2020-02-17 | 2023-10-03 | 東芝ライテック株式会社 | Cap inspection device and capsule inspection method |
IT202000015469A1 (en) * | 2020-06-26 | 2021-12-26 | Metacortex Srl | SYSTEM AND METHOD FOR MEASURING THE PHYSICAL CHARACTERISTICS OF PELLETS AND DEVICE PROVIDED IN THE SYSTEM |
KR102598333B1 (en) * | 2021-03-25 | 2023-11-06 | 재단법인대구경북과학기술원 | System for inspecting double-sided of inspection object using a plurality of conveyors and method thereof |
CN117607143B (en) * | 2023-10-16 | 2024-06-25 | 江苏跃畅新能源科技有限公司 | Device and method for detecting separation efficiency of solid broken mixed materials |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0647239A (en) * | 1992-07-31 | 1994-02-22 | Matsushita Electric Ind Co Ltd | Humidifying and dehumidifying machine |
JP2002062113A (en) * | 2000-08-17 | 2002-02-28 | Ishii Ind Co Ltd | Method and device for measuring object to be detected |
JP2005083775A (en) * | 2003-09-05 | 2005-03-31 | Seirei Ind Co Ltd | Grain classifier |
JP2008020347A (en) * | 2006-07-13 | 2008-01-31 | Akita Prefecture | Legume discrimination structure |
CN103344575A (en) * | 2013-07-01 | 2013-10-09 | 江南大学 | Hyperspectral-image-technology-based multi-quality nondestructive testing method for dried green soybeans |
TW201415005A (en) * | 2012-06-27 | 2014-04-16 | Satake Eng Co Ltd | Grain external appearance measurement device |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3371915B2 (en) * | 1993-06-04 | 2003-01-27 | 株式会社エフ・エフ・シー | Appearance inspection method |
JP4235776B2 (en) * | 1998-12-21 | 2009-03-11 | 株式会社サタケ | Granular color sorter |
JP5560431B2 (en) * | 2008-02-13 | 2014-07-30 | 秋田県 | Fruit discrimination structure |
JP2012103092A (en) * | 2010-11-10 | 2012-05-31 | Seiko Epson Corp | Reading device |
JP2013094764A (en) * | 2011-11-04 | 2013-05-20 | Yamamoto Co Ltd | Pod sorting apparatus |
-
2016
- 2016-02-18 TW TW105104823A patent/TWI632360B/en active
- 2016-02-18 JP JP2017500739A patent/JP6590907B2/en active Active
- 2016-02-18 WO PCT/JP2016/054762 patent/WO2016133175A1/en active Application Filing
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0647239A (en) * | 1992-07-31 | 1994-02-22 | Matsushita Electric Ind Co Ltd | Humidifying and dehumidifying machine |
JP2002062113A (en) * | 2000-08-17 | 2002-02-28 | Ishii Ind Co Ltd | Method and device for measuring object to be detected |
JP2005083775A (en) * | 2003-09-05 | 2005-03-31 | Seirei Ind Co Ltd | Grain classifier |
JP2008020347A (en) * | 2006-07-13 | 2008-01-31 | Akita Prefecture | Legume discrimination structure |
TW201415005A (en) * | 2012-06-27 | 2014-04-16 | Satake Eng Co Ltd | Grain external appearance measurement device |
CN103344575A (en) * | 2013-07-01 | 2013-10-09 | 江南大学 | Hyperspectral-image-technology-based multi-quality nondestructive testing method for dried green soybeans |
Also Published As
Publication number | Publication date |
---|---|
TW201636596A (en) | 2016-10-16 |
WO2016133175A1 (en) | 2016-08-25 |
JPWO2016133175A1 (en) | 2018-01-25 |
JP6590907B2 (en) | 2019-10-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
TWI632360B (en) | Pod screening system and pod screening device | |
US10638730B2 (en) | Imaging for determination of crustacean physical attributes | |
Mahendran et al. | Application of computer vision technique on sorting and grading of fruits and vegetables | |
KR101298957B1 (en) | Wood knot detecting method, device, and program | |
US9129350B2 (en) | Systems and methods to analyze an immunoassay test strip comb member | |
JP5213038B2 (en) | Strawberry quality measuring method and strawberry quality measuring device | |
JP4590553B2 (en) | Nondestructive judgment method for ginger damaged grains | |
KR102037560B1 (en) | Device for measuring appearance of grains | |
CN111353992A (en) | Agricultural product defect detection method and system based on textural features | |
CN113390799B (en) | Method for identifying and detecting stems in tobacco leaves | |
JPH0534281A (en) | Evaluating apparatus for appearance of melon | |
Noordam et al. | Detection and classification of latent defects and diseases on raw French fries with multispectral imaging | |
US11636588B2 (en) | Apparatus and system for assessing paddy rice grains | |
JP2002005842A (en) | Non-contact detector | |
JP2016075660A (en) | Boiled egg quality determination device | |
JPH11230960A (en) | Automatic judgment apparatus for freshness of egg | |
JP2003251282A (en) | System for measuring and deciding grade and class of strawberry | |
JP6948032B2 (en) | Image analyzer and inspection system | |
Visen | Machine vision based grain handling system | |
Nagata et al. | Quality Evaluation of | |
JP2022129666A (en) | Legume selection system and legume selection device | |
JP2022129668A (en) | Legume selection system and legume selection device | |
JP2022129667A (en) | Legume selection system and legume selection device | |
Throop et al. | Multispectral inspection station detects defects on apples | |
RU2624705C1 (en) | Method for determining anatomical and morphological defects of grain and seeds in grain crops batches |