JPH04184573A - Visual device, such as fruit collecting machine or the like - Google Patents

Visual device, such as fruit collecting machine or the like

Info

Publication number
JPH04184573A
JPH04184573A JP2314997A JP31499790A JPH04184573A JP H04184573 A JPH04184573 A JP H04184573A JP 2314997 A JP2314997 A JP 2314997A JP 31499790 A JP31499790 A JP 31499790A JP H04184573 A JPH04184573 A JP H04184573A
Authority
JP
Japan
Prior art keywords
luminance
objects
hue
light
plural
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP2314997A
Other languages
Japanese (ja)
Inventor
Hidetaka Hirayama
秀孝 平山
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Iseki and Co Ltd
Iseki Agricultural Machinery Mfg Co Ltd
Original Assignee
Iseki and Co Ltd
Iseki Agricultural Machinery Mfg Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Iseki and Co Ltd, Iseki Agricultural Machinery Mfg Co Ltd filed Critical Iseki and Co Ltd
Priority to JP2314997A priority Critical patent/JPH04184573A/en
Publication of JPH04184573A publication Critical patent/JPH04184573A/en
Pending legal-status Critical Current

Links

Landscapes

  • Closed-Circuit Television Systems (AREA)
  • Control Of Position Or Direction (AREA)
  • Harvesting Machines For Specific Crops (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Geophysics And Detection Of Objects (AREA)
  • Image Processing (AREA)
  • Transforming Light Signals Into Electric Signals (AREA)

Abstract

PURPOSE:To individually discriminate plural objects, such as fruits, etc., overlapping each other in front and in rear in a short time by finding the AND of a hue component analyzed result and luminance component analyzed result. CONSTITUTION:The outline of the whole body of a group of plural objects overlapping each other in front and in rear is detected by reducing a picture to primary color components by means of a hue analyzing means (image sensor camera) 2 and binarizing the primary color components into a specific hue component and the other hue components by means of a video interface 3. In addition, since a luminance analyzing means analyzes the luminance of the objects, the boundary section of the objects on the rear side is detected as a low-luminance belt when the plural objects overlap each other in front and in rear. Such low-luminance belt is produced, since the objects are irradiated with rays of light (light from a light source 10 and natural light) arriving at the objects in different directions from plural light sources and shades formed at the boundary section lower the luminance. The AND of the detected results of the hue and luminance analyzing means is analyzed by means of an analyzing means (CPU). Therefore, the objects can be individually discriminated quickly and, at the same time, the distance from this device to each object can also be discriminated.

Description

【発明の詳細な説明】 [産業上の利用分野] 本発明は、無人の果実収穫機等に設けられる視覚装置に
関するものである。
DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a visual device installed in an unmanned fruit harvesting machine or the like.

[従来の技術] トマト等の作物が植え付けられている畝に沿って自走し
つつ、収穫物探索用の視覚装置で成熟した果実を探し出
し、これを遠隔操作される収穫ノ\ントで収穫するよう
にした無人の果実収穫機が開発されている。
[Conventional technology] While moving along the ridge where crops such as tomatoes are planted, a visual device for searching for crops locates mature fruits, which are then harvested using a remotely controlled harvesting node. An unmanned fruit harvesting machine has been developed.

上記視覚装置は、イメージセンサカメラが撮った画像を
所定の色相成分(対象物がトマトである場合は赤)とそ
れ以外の色相成分(例えば青および緑)とに二値化する
ことにより果実を識別するようにしたものが一般的で、
該メージセンサカメラを随時移動させて収穫可能な果実
を探索するようになっている。
The visual device described above binarizes the image taken by the image sensor camera into a predetermined hue component (red if the object is a tomato) and other hue components (for example, blue and green). It is common to identify
The image sensor camera is moved at any time to search for fruits that can be harvested.

[発明が解決しようとする課!l] しかしながら1例えば第2図に示すように複数側の果実
T(A、B、C)が重なり合って存在する場合、上記従
来の視覚装置によれば、第3図に示す如く果実群の輪瑯
だけが認識されるので、何個の果実が存在するのか、果
実の中心はどこか、また相互の遠近関係はどのようなっ
ているのかを瞬時に判定できなかった。そこで従来は、
数学的な演算を用いて複数果実の個別化を行なっていた
が、この方法は、処理に時間を要するため、迅速な対応
ができないという問題があった。
[The problem that the invention tries to solve! However, if there are fruits T (A, B, C) on a plurality of sides overlapping each other as shown in FIG. Since only the sacs were recognized, it was not possible to instantly determine how many fruits were present, where the center of the fruit was, or what their relative distances were. Therefore, conventionally,
Previously, multiple fruits were separated using mathematical operations, but this method required time to process, so there was a problem in that it was not possible to respond quickly.

[課題を解決するための手段] と記課題を解決するために、本発明は次のような構成と
した。
[Means for Solving the Problems] In order to solve the problems described above, the present invention has the following configuration.

すなわち1本発明にかかる視覚装置は、照射方向が異な
る複数の光源から対象物に投光する投光手段と、対象物
からの反射光を撮像するイメージセンサカメラと、該イ
メージセンサカメラの画像を原色成分に分析する色相分
析手段と、同じく輝度成分に分析する輝度分析手段と、
前記色相分析手段および輝度分析手段による分析結果の
論理積より、互いに前後に重なり合って存在する複数の
対象物の位置関係を解析する解析手段とを備えてなる。
In other words, the visual device according to the present invention includes: a light projector that projects light onto an object from a plurality of light sources with different irradiation directions; an image sensor camera that captures reflected light from the object; and an image sensor camera that captures an image of the image sensor camera. A hue analysis means for analyzing into primary color components, a luminance analysis means for analyzing into luminance components,
and analysis means for analyzing the positional relationship of a plurality of objects existing one after another, based on the logical product of the analysis results obtained by the hue analysis means and the brightness analysis means.

[作 用] 色相分析手段で画像を原色成分に分析し、これを特定の
色相成分とそれ以外の色相成分とに二値化することによ
り、複数の重なり合った対象物群全体の輪郭が検出され
る。また、輝度分析手段で対象物の輝度分析を行なうこ
とにより、複数の対象物が重なり合っている場合には、
後方に位置する対象物の境界部が輝度の低い帯状に検出
される。
[Operation] By analyzing an image into primary color components using a hue analysis means and binarizing this into a specific hue component and other hue components, the outline of the entire group of multiple overlapping objects can be detected. Ru. In addition, by analyzing the brightness of objects using a brightness analysis means, if multiple objects overlap,
The boundary of the object located at the rear is detected as a band with low brightness.

これは、照射方向の異なる複数の光源から投光されてい
るので、上記部分に影が生じ輝度が低くなっているから
である。これら色相分析手段と輝度分析手段の検出結果
の論理積を解析手段で解析することにより、各々の対象
物が個別に識別されるとともに、それらの遠近関係が判
定される。
This is because light is emitted from a plurality of light sources with different irradiation directions, so a shadow is created in the above-mentioned area and the brightness is low. By analyzing the logical product of the detection results of the hue analysis means and the brightness analysis means by the analysis means, each object is individually identified and the distance relationship between them is determined.

[実施例] 以下、本発明の1例として第20図に示すトマト用の果
実収穫機30の視覚装置について説明する。なお、図中
の31は収穫ハンド31.32は収穫物容器で、視覚装
z1は収穫ハンド31に取り付けられている。
[Example] Hereinafter, as an example of the present invention, a visual device of a tomato fruit harvester 30 shown in FIG. 20 will be described. In addition, 31 in the figure is a harvesting hand 31, 32 is a harvest container, and the visual device z1 is attached to the harvesting hand 31.

この視覚装Mlは1作用ごとに便宜上分離して表記した
第1図、第5図および第9図であられされ、イメージセ
ンサカメラ(CCD)2の画像をビデオインターフェイ
ス3で水工走査し、抽出された原色信号R(赤)、G(
緑)、B(i)と輝度信号Yを処理して対象物(例えば
トマト)を知覚識別するようになっている0画像処理に
あた1ては第1図に示す個別認識処理、第5図に示すサ
イズ判定処理、および第9図に示すノイズ除去処理が行
なわれる。なお、これら各図においてR−YおよびG−
Yは、赤または緑の原色信号から一度信号を差し引いた
もの、DAI〜3は予め設定されている数値、H,5Y
NCは走査線が画面を水平に1列走査し終わるごとに出
される信号、V+ 5YNCは走査線が1画面の下段ま
で操作し終わるごとに出される信号である。以下、これ
ら各処理について順に説明する。
This visual system Ml is shown in FIGS. 1, 5, and 9, which are shown separately for each action for convenience, and is extracted by scanning the image of an image sensor camera (CCD) 2 with a video interface 3. The primary color signals R (red), G (
For image processing, 1 is the individual recognition processing shown in FIG. 1, and 5 The size determination process shown in the figure and the noise removal process shown in FIG. 9 are performed. In addition, in each of these figures, R-Y and G-
Y is the signal once subtracted from the red or green primary color signal, DAI ~ 3 is a preset value, H, 5Y
NC is a signal that is output every time the scanning line finishes scanning one row horizontally on the screen, and V+5YNC is a signal that is output every time the scanning line finishes scanning the bottom row of one screen. Each of these processes will be explained in turn below.

個別識別処理は1前後に重なり合った複数の対象物を各
々個別に識別し、その前後関係を判定するための処理で
ある。第2図に示すように複数の果実T(A、B、C)
が前後に重なり合って存在している場合、これを色相成
分によって二値化しただけでは第3図のようにl為され
る。そこで、照射方向の異なる少なくとも二つ以−Eの
光源から対象物に投光し1色組成分による二値化とは別
に輝度成分による二値化を行ない、これら二つのデータ
の論理積を演算することにより、14fNに示すように
認識される。すなわち、果実と果実の境界部の影の部分
が「0」と判定され、個々の果実ごとに分離された状態
で認識されるのである。したがって、複数の果実の個別
化が容易に行なわれる。また1手前側に位置する果実は
ど隠れている部分が少ないため大きく認識されるので、
後記サイズ判定回路の判定結果に基づき、−度の収穫動
作ごとに一番大きいを収穫するようにすればよい。
The individual identification process is a process for individually identifying a plurality of objects that overlap one another, and determining the context of the objects. Multiple fruits T (A, B, C) as shown in Figure 2
When there are overlapping images one after the other, simply binarizing them using hue components will result in the result as shown in FIG. 3. Therefore, light is emitted onto the object from at least two or more light sources with different irradiation directions, and in addition to binarization based on one color composition, binarization is performed based on the luminance component, and the logical product of these two data is calculated. By doing so, it is recognized as shown in 14fN. That is, the shaded portion at the boundary between fruits is determined to be "0", and each fruit is recognized separately. Therefore, individualization of a plurality of fruits is easily performed. Also, the fruit located one step closer to you will be recognized larger because there is less hidden part.
Based on the determination result of the size determination circuit described later, the largest size may be harvested every - degree of harvesting operation.

昼間に収穫作業を行なう場合は、ハウス内に自然光が照
射しているので、投光手段として別に1個の光源(例え
ばストロボランプ)lOを用意すればよい、自然光の照
射方向、果実収穫機の進行方向等に応じて、光源移動用
アクチュエータ11で光源IOの照射方向を適宜調整す
る。
When harvesting is done during the day, natural light shines into the greenhouse, so you only need to prepare one additional light source (for example, a strobe lamp) as a lighting means. The light source moving actuator 11 adjusts the irradiation direction of the light source IO as appropriate depending on the traveling direction and the like.

サイズ判定処理は、二値化によって対象物の輪郭が判明
した画像データより、その対象物の大きさをリアルタイ
ムで検出するための処理である。
The size determination process is a process for detecting the size of an object in real time from image data whose outline has been determined by binarization.

1例として第6図に示す果実Tの画像を処理する場合に
ついて説明する。まず、第7図に示す如く。
As an example, a case will be described in which an image of a fruit T shown in FIG. 6 is processed. First, as shown in FIG.

各未刊走査ごとに果実と背景との境界部に位置する2つ
の点(画、i)U+、U2をアドレスカウンタ13で記
憶し、その中間点Ucを演算窓14で演算しておく、そ
して、各水平走査を終了したのち上記中間点Uc  、
・・・を結ぶと、これが第8図に示すような1本の線り
として認識される。この線りの長さが果実の縦方向の長
さをあられしている。
For each unprinted scan, two points (pictures, i) U+ and U2 located at the boundary between the fruit and the background are stored in the address counter 13, and the intermediate point Uc is calculated in the calculation window 14, and After completing each horizontal scan, the intermediate point Uc,
When connected, this is recognized as a single line as shown in Figure 8. The length of this line determines the vertical length of the fruit.

さらに、この線りの中心Oが当該果実の中心をあられし
ている。このように2点の中間点を求めるという筒型な
演算処理を行なうだけで対象物の太Jさと中心を求める
ことができるため、処理が迅速に行なわれ、果実の大き
さおよび中心をリアルタイムで検出することができるの
である。
Furthermore, the center O of this line marks the center of the fruit. In this way, the thickness and center of the object can be determined simply by performing the cylindrical calculation process of finding the midpoint between two points, so the processing is quick and the size and center of the fruit can be determined in real time. It can be detected.

実った状態にある果実には光のよく当る部分や影の部分
があり、これらの影響で画像処理する上で不都合なノイ
ズが生じることがある。このノイズを除去(フィルタリ
ング)するのがノイズ除去処理である0例えば、第10
図に示す対象物を赤い成分とそれ以外の色相の成分に二
値化すると第11図のようになる。第11図において、
トマトTのハイライト部20および暗部21と葉22゜
・・・の部分にノイズN、−・・が発生しているのが見
られる8画像処理にあたっては、画素24.・・・ごと
の主色相を判定し、赤と判定し、た画素にrlJ、赤で
ないと判定した画素にrQJを入力する。これを模式的
にあられすと第12図のようになり、トマトの内部に赤
く認識されない画素a、−・・が存在したり、トマトの
外に赤く認識された画素す。
A ripe fruit has parts that are often exposed to light and parts that are in shadow, and these effects can cause noise that is inconvenient during image processing. Removing (filtering) this noise is the noise removal process.0For example, the 10th
When the object shown in the figure is binarized into a red component and other hue components, the result is as shown in FIG. 11. In Figure 11,
Noise N, -... can be seen occurring in the highlight part 20, dark part 21, and leaf 22° of tomato T. 8 In processing the image, the pixel 24. . . . rlJ is input to the pixel determined to be red, and rQJ is input to the pixel determined not to be red. If this happens schematically, it will look like Figure 12, where there are pixels a, ... that are not recognized as red inside the tomato, and pixels that are recognized as red outside of the tomato.

・・・、Cが存在し、正確なトマトの形状を把握できな
くなる。このよう場合、収穫ハンドが誤動作を行なうお
それがある。フィルタリングの方法は、水平に連続する
所定数の画素24.・・・を比較することにより1画素
ごとにノイズであるか否かを判定し、ノイズでないと判
定したときはデジタルコンパレータ■16によりフリッ
プフロップ17を「セット■」、ノイズであると判定し
たときはデジタルコンパレータ■18によりフリップフ
ロップ17を「リセ−y)(DJする0例えば、フィル
タリングレベル「4」は4画素を比較するもので、横方
向に並ぶ連続する4つの画素が全て「0」のときは始点
の画XをrQJ、4つの画素が全て「1」のと5は始点
の画1t−rlJとし、4つの画素に「0」とrlJと
が混在しているときは始点の画素を左隣りの画素と同一
とする。このような処理を走査順に逐次施すと、画像が
第13図のように修正される。修正後の画像は、ノイズ
の多く(aとb)が除去されており、トマトの形状が明
瞭に認識されるようになる。なお、ノイズの発生場所に
よっては1例えばノイズCのようにノイズでないと誤認
識される場合があるが、それによる影響は小さく、無視
することができる。
..., C exists, making it impossible to grasp the exact shape of the tomato. In such a case, the harvesting hand may malfunction. The filtering method uses a predetermined number of horizontally consecutive pixels 24. By comparing ..., it is determined whether each pixel is noise or not. When it is determined that it is not noise, the flip-flop 17 is set by the digital comparator ■ 16, and when it is determined that it is noise. For example, filtering level "4" compares four pixels, and if all four consecutive pixels lined up in the horizontal direction are "0", When all four pixels are "1", the starting point pixel It is assumed to be the same as the pixel on the left.If such processing is performed sequentially in the scanning order, the image is modified as shown in Figure 13.The modified image has much of the noise (a and b) removed. As a result, the shape of the tomato can be clearly recognized.Note that depending on where the noise occurs, it may be mistakenly recognized as not noise, such as Noise C, but the effect of this is small and can be ignored. be able to.

フィルタリングレベルは、第14図に示すように、特定
の色相信号(対象物がトマトの場合は赤色信号R)で特
徴づけられる画素群全体の平均輝度に応じて適正最小値
が決定される。フィルタリングレベルが低い方が処理を
迅速に行なえるので好ましいが、フィルタリングレベル
を下げすぎると、平均輝度が小さい場合にトマトの一部
分をトマトでないと判定したり、平均輝度が大きい場合
にトマトでない部分もトマトであると判定してしまうお
それがある。
For the filtering level, as shown in FIG. 14, an appropriate minimum value is determined according to the average brightness of the entire pixel group characterized by a specific hue signal (red signal R when the object is a tomato). A low filtering level is preferable because processing can be done quickly, but if the filtering level is lowered too much, parts of the tomato may be determined to be not tomatoes when the average brightness is low, or parts that are not tomatoes may be determined to be non-tomatoes when the average brightness is high. There is a risk that it will be determined to be a tomato.

つぎに、第15図は中心判定処理の異なるX篇例をあら
れし、この中心判定処理法によれば、第16図に示すよ
うに、対象物画像における左端(もしくは下端)の画J
Uu  (x+  、 V+ )と左端(もしくは右端
)の画素Ul  (X2  、72 )から対象物の中
心位置Uo(x+  、 yz )を求めるようになっ
ている。このように構成すると各水モ走査線ごとに中心
を求める必要がないので、前記第5図に示したサイズ判
定処理法と比較しても、さらに対象物の中心の検出を高
速化することがきる。
Next, FIG. 15 shows X examples of different center determination processing. According to this center determination processing method, as shown in FIG.
The center position Uo (x+, yz) of the object is determined from Uu (x+, V+) and the leftmost (or rightmost) pixel Ul (X2, 72). With this configuration, there is no need to find the center for each water mop scanning line, so even compared to the size determination processing method shown in FIG. 5 above, the detection of the center of the object can be made faster. Wear.

また、第17図はさらに異なる中心判定処理法をあられ
している。外方に凸で曲面状の滑らかな外周面を有する
物体であるといラドマドの特性により、第18図に示す
ように光の照射方向と同軸上から見た場合、トマトの中
心部の輝度が最も高くなる。そこで、CCD2の基線と
同軸上に光源10を設け、第19図の如く、各水平走査
ごとに最高輝度の点Ubを検出するとともに、ぞれをビ
−クホールl” l 9に記憶し、ざらに1画面の走査
終了後これら各走査線ごとのUb、・・・の中で最も明
るい点Ub■axを検出し、このUbmxをトマ)Tの
中心とするのである。このように、この中心判定処理法
は、各画素の輝度を比較するだけでリアルタイムで果実
の中心が求められる。
Further, FIG. 17 shows a further different center determination processing method. Due to the characteristics of Radoma, which is an object with an outwardly convex, curved, smooth outer peripheral surface, when viewed from the same axis as the direction of light irradiation, as shown in Figure 18, the brightness at the center of the tomato is the highest. It gets expensive. Therefore, a light source 10 is installed on the same axis as the base line of the CCD 2, and as shown in FIG. After scanning one screen, the brightest point Ubax is detected among these scanning lines Ub, . . . and this Ubmx is set as the center of T.In this way, this center The determination processing method determines the center of the fruit in real time by simply comparing the brightness of each pixel.

[発明の効果] 以上に説明したように、本発明にかかる視覚装置は、色
相成分による分析と輝度成分による分析の論理積を求め
ることにより、例えば果実等の前後に重なって存在する
複数の対量物を短時間で個別に識別することが可能とな
った。
[Effects of the Invention] As explained above, the visual device according to the present invention calculates the logical product of the analysis based on the hue component and the analysis based on the luminance component. It has become possible to identify individual objects in a short time.

【図面の簡単な説明】[Brief explanation of drawings]

第1図、第5図および第9図は本発明の1例である視覚
装置のブロック図、第15図および第17図はともに異
なる視覚装置のブロック図、第2図乃至第4図、第6図
乃至第8図、第10図乃至第13図、第16図、第18
図および第19図はいずれも画像処理の説明図、第14
図は平均−度とフィルタリングレベルの関係を示す図、
第20図は果実収穫機の1例の側面図である。 1・・・視覚装置、2−・・イメージセンサカメラ、3
・・・ビデオインターフェイス、10・・・光源、30
・−・果実収穫機。
1, 5 and 9 are block diagrams of a visual device which is an example of the present invention, FIGS. 15 and 17 are block diagrams of different visual devices, and FIGS. 2 to 4, Figures 6 to 8, Figures 10 to 13, Figures 16, 18
Both Fig. 19 and Fig. 19 are explanatory diagrams of image processing.
The figure shows the relationship between average degree and filtering level.
FIG. 20 is a side view of an example of a fruit harvesting machine. 1...Visual device, 2-...Image sensor camera, 3
...Video interface, 10...Light source, 30
・−・Fruit harvesting machine.

Claims (1)

【特許請求の範囲】[Claims] (1)照射方向が異なる複数の光源から対象物に投光す
る投光手段と、対象物からの反射光を撮像するイメージ
センサカメラと、該イメージセンサカメラの画像を原色
成分に分析する色相分析手段と、同じく輝度成分に分析
する輝度分析手段と、前記色相分析手段および輝度分析
手段による分析結果の論理積より、互いに前後に重なり
合って存在する複数の対象物の位置関係を解析する解析
手段とを備えてなる視覚装置。
(1) A light projector that projects light onto an object from a plurality of light sources with different irradiation directions, an image sensor camera that captures the reflected light from the object, and a hue analysis that analyzes the image of the image sensor camera into primary color components. a luminance analysis means that similarly analyzes the luminance components; and an analysis means that analyzes the positional relationship of a plurality of objects that overlap one another based on the logical product of the analysis results of the hue analysis means and the luminance analysis means. A visual device equipped with
JP2314997A 1990-11-19 1990-11-19 Visual device, such as fruit collecting machine or the like Pending JPH04184573A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2314997A JPH04184573A (en) 1990-11-19 1990-11-19 Visual device, such as fruit collecting machine or the like

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2314997A JPH04184573A (en) 1990-11-19 1990-11-19 Visual device, such as fruit collecting machine or the like

Publications (1)

Publication Number Publication Date
JPH04184573A true JPH04184573A (en) 1992-07-01

Family

ID=18060170

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2314997A Pending JPH04184573A (en) 1990-11-19 1990-11-19 Visual device, such as fruit collecting machine or the like

Country Status (1)

Country Link
JP (1) JPH04184573A (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
ES2296452A1 (en) * 2005-06-29 2008-04-16 Universidad Politecnica De Madrid Detector for vegetable stem and stalk, has laser line that is diode laser spot with conventional lens that spreads light plane at angle to produce mark, and images of scene are taken by vision sensor
JP2011191956A (en) * 2010-03-12 2011-09-29 Japan Research Institute Ltd Farm product purchase reservation method, farm product purchase reservation system, and farm product purchase reservation device
JP2011192022A (en) * 2010-03-15 2011-09-29 Japan Research Institute Ltd Farm product monitoring method, farm product monitoring system, and farm product monitoring device
JP2011197832A (en) * 2010-03-17 2011-10-06 Japan Research Institute Ltd Field crop sales method, field crop sales system and field crop sales apparatus
CN109744559A (en) * 2017-11-07 2019-05-14 财团法人资讯工业策进会 Fruit accelerates the ripening system and fruit forced ripening method
JP2019083750A (en) * 2017-11-07 2019-06-06 シブヤ精機株式会社 Fruit/vegetable harvesting device
JP2020195335A (en) * 2019-06-04 2020-12-10 本田技研工業株式会社 Fruit vegetable harvesting device and fruit vegetable harvesting method

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
ES2296452A1 (en) * 2005-06-29 2008-04-16 Universidad Politecnica De Madrid Detector for vegetable stem and stalk, has laser line that is diode laser spot with conventional lens that spreads light plane at angle to produce mark, and images of scene are taken by vision sensor
JP2011191956A (en) * 2010-03-12 2011-09-29 Japan Research Institute Ltd Farm product purchase reservation method, farm product purchase reservation system, and farm product purchase reservation device
JP2011192022A (en) * 2010-03-15 2011-09-29 Japan Research Institute Ltd Farm product monitoring method, farm product monitoring system, and farm product monitoring device
JP2011197832A (en) * 2010-03-17 2011-10-06 Japan Research Institute Ltd Field crop sales method, field crop sales system and field crop sales apparatus
CN109744559A (en) * 2017-11-07 2019-05-14 财团法人资讯工业策进会 Fruit accelerates the ripening system and fruit forced ripening method
JP2019083750A (en) * 2017-11-07 2019-06-06 シブヤ精機株式会社 Fruit/vegetable harvesting device
JP2020195335A (en) * 2019-06-04 2020-12-10 本田技研工業株式会社 Fruit vegetable harvesting device and fruit vegetable harvesting method

Similar Documents

Publication Publication Date Title
Bulanon et al. Development of a real-time machine vision system for the apple harvesting robot
EP3801934B1 (en) Process and system for in-line inspection of product stream for detection of foreign objects
CN111239142A (en) Paste appearance defect detection device and method
CN111583193A (en) Pistachio nut framework extraction device based on geometric contour template matching and algorithm thereof
JPH04184573A (en) Visual device, such as fruit collecting machine or the like
JPH05284501A (en) Intrusion monitor picture device
JP3614980B2 (en) Agricultural product appearance inspection method and apparatus
JPH0534281A (en) Evaluating apparatus for appearance of melon
US10909713B2 (en) System and method for item location, delineation, and measurement
JPH0554124A (en) Visual device for fruit harvesting robot or the like
JP3729562B2 (en) Automatic monitoring device
JPH03165737A (en) Detecting method for line of sight direction
JPH0658733A (en) Inspecting method of nonuniform section of glass bottle
JPH10137696A (en) Separation method and its device
JP3506161B2 (en) Agricultural and fishery product appearance inspection device
JP2001116534A (en) Circularity deciding method, circularity arithmetic unit, and recording medium
JPH09163853A (en) Image processing device for fruit
Patel et al. Monochrome computer vision for detecting quality defects of fruits and vegetables
JP4656399B2 (en) Agricultural products visual inspection equipment
JPH02218944A (en) Method and apparatus for detecting white flaw of citrus fruit
JPH08166380A (en) Automatic bruise detector for vegitable and fruit
Lu et al. Obstacle recognition using multi-spectral imaging for citrus picking robot
JP2949736B2 (en) Fruit detection method
JPS61286986A (en) Fruit recognizing device
JP2696166B2 (en) Fruit and vegetable testing device