TWM583944U - A rotary-disk-based system for coffee bean sorting - Google Patents
A rotary-disk-based system for coffee bean sorting Download PDFInfo
- Publication number
- TWM583944U TWM583944U TW108206223U TW108206223U TWM583944U TW M583944 U TWM583944 U TW M583944U TW 108206223 U TW108206223 U TW 108206223U TW 108206223 U TW108206223 U TW 108206223U TW M583944 U TWM583944 U TW M583944U
- Authority
- TW
- Taiwan
- Prior art keywords
- coffee
- coffee beans
- turntable
- coffee bean
- screening system
- Prior art date
Links
Abstract
Description
本創作係關於咖啡豆篩選系統，尤指一種具有轉盤的咖啡豆篩選系統。 This creation is about coffee bean screening systems, especially a coffee bean screening system with a turntable.
查，諸多醫學期刊指出，咖啡中尚有多種對人體健康有益的成分，如咖啡因，其能激活中樞神經系統，且能抵抗困倦，降低傷風、感冒的機率，並減緩哮喘與水腫的發生；如抗氧化物質(antioxidants)，其能延緩肝疾惡化、減少慢性肝臟病流行率及降低肝硬化併發症的死亡風險；如抗癡呆物質，其能減少有害物對身體的影響，並降低人體腦中導致失憶的類澱粉含量；如多酚化合物，其能延緩低密度脂蛋白氧化，並溶解血液凝塊及防止血栓；因此，隨著咖啡的好處被一一揭露，亦使得飲用咖啡的人口逐漸上升，造成咖啡文化的興起。 Cha, many medical journals pointed out that there are many kinds of ingredients that are beneficial to human health, such as caffeine, which can activate the central nervous system, resist drowsiness, reduce the chance of colds and colds, and slow the occurrence of asthma and edema; Such as antioxidants, which can delay the deterioration of liver disease, reduce the prevalence of chronic liver disease and reduce the risk of death from cirrhosis complications; such as anti-dementia substances, which can reduce the impact of harmful substances on the body and reduce the human brain The starch-like content that causes amnesia; such as polyphenolic compounds, which delay the oxidation of low-density lipoproteins, dissolve blood clots and prevent blood clots; therefore, as the benefits of coffee are revealed one by one, the population of coffee is gradually Rising, causing the rise of coffee culture.
一般來說，為了保持咖啡烘焙後的風味與品質，目前咖啡豆的製作過程中，普遍具有「分級」與「篩選」等步驟，其中，「分級」是將咖啡豆按外觀大小分成不同等級，使各級咖啡豆有其一致性，以達到增加產品價值，且有助於後續烘焙時，保持咖啡豆品質的一致性；又，「篩選」則是要將其中的異物與瑕疵豆挑出，前述異物包含了石頭、木屑、土粒等 非咖啡豆的外來物質，前述瑕疵豆則能根據SCAA美國精品咖啡協會所列舉的黑豆(Black)、酸豆(Sour)、咖啡櫻桃豆筴(Dried Cherry/Pod)、發霉豆(Fungus Damage)、蟲蛀豆(Insect Damage)、破裂豆(Broken)、未熟豆(Immature)、萎縮豆(Withered)、貝殼狀(Shell)、漂浮豆(Floater)、羊皮紙(parchment)、豆殼(Hull)、奎克豆(Quakers)等，畢竟，當所販售的咖啡豆包含瑕疵豆時，不但會影響咖啡風味，嚴重時更會對人體造成傷害，例如發霉豆所產生的黃麴毒素。 In general, in order to maintain the flavor and quality of coffee after baking, the current process of making coffee beans generally has the steps of "grading" and "screening". Among them, "grading" is to divide the coffee beans into different grades according to the size of the appearance. Make the coffee beans at all levels consistent, in order to increase the value of the product, and to help maintain the consistency of the quality of the coffee beans during the subsequent baking; in addition, the "screening" is to pick out the foreign bodies and the cowpeas. The aforementioned foreign matter includes stones, wood chips, soil particles, and the like. For foreign substances other than coffee beans, the aforementioned cowpeas can be based on Black, Sour, Dried Cherry/Pod, Fungus Damage, etc. listed by the SCAA American Fine Coffee Association. Insect Damage, Broken, Immature, Withered, Shell, Floater, Parchment, Hull, Kui Quakers, etc. After all, when the coffee beans sold contain cowpea, it will not only affect the flavor of the coffee, but also cause harm to the human body in severe cases, such as the toxin produced by the moldy beans.
目前咖啡豆的篩選方式，除了以人工挑選之外，尚會採用機器作為輔助，舉例來說，有業者會選擇使用比重選豆機，並利用風力或震動的方式，使咖啡豆能依照其顆粒大小與重量進行分類，然而，比重選豆機的篩選方式僅能夠進行初步分類，無法有效篩選出顏色上的瑕疵，例如：局部發霉、黑豆...等。為能解決前述問題，有業者會選擇使用色選機，以能依照咖啡豆的顏色來篩選異物與瑕疵豆，如習知色選機(如台灣第375537號專利案)即是在咖啡豆掉落的過程中，擷取咖啡豆的影像，以進行辨識並同時剔除其中異物與瑕疵豆，但是，由於每顆咖啡豆的重量不同，造成其掉落時間點有差異，難以準確抓住剔除時機，且掉落過程中，時常發生複數顆咖啡豆彼此間相互遮蔽，造成錯誤判斷之情事，導致篩選結果不佳。 At present, the screening method of coffee beans, in addition to manual selection, will use the machine as an aid. For example, some operators will choose to use the specific gravity to select the bean machine, and use wind or vibration to make the coffee beans according to their particles. The size and weight are classified. However, the screening method of the specific gravity machine can only be used for preliminary classification, and it is impossible to effectively screen out the defects on the color, for example, local mold, black beans, and the like. In order to solve the above problems, some operators will choose to use a color sorter to screen foreign bodies and cowpeas according to the color of the coffee beans. For example, the conventional color sorting machine (such as Taiwan Patent No. 375537) is in the coffee beans. In the process of falling, the image of the coffee beans is taken to identify and simultaneously remove the foreign matter and the cowpea. However, because the weight of each coffee bean is different, the time of the drop is different, and it is difficult to accurately grasp the timing of the rejection. In the process of falling, it is often the case that a plurality of coffee beans are shielded from each other, causing misjudgment, resulting in poor screening results.
除了前述在掉落過程中進行篩選的色選機之外，如台灣第M570428號專利案尚提出了另一種色選機，其會在震動盤上設有透明的螺旋斜面，並藉由震動盤的震動效果，令咖啡豆能夠被推擠至螺旋斜面上，以被擷取影像並進行篩選。前述色選機雖解決了在掉落過程中進行篩選所會衍生的問題，但是，創作人發現，在實施使用上，前述色選機仍有諸多缺 失，首先，震動盤普遍為金屬材質製成，因此，要在其上額外以其它透明材質一併製作，在工藝上極為複雜，且難以商品化；其次，震動盤在震動輸送的過程中，是藉由其震動推送來運送咖啡豆，因此，實際上仍容易發生咖啡豆堆疊的情況，尤其是，當螺旋斜面較為窄小時更容易產生前述情事，影響擷取影像品質。 In addition to the aforementioned color sorting machine which is screened during the falling process, another color sorting machine is proposed in the patent No. M570428 of Taiwan, which is provided with a transparent spiral bevel on the vibrating plate and is provided with a vibrating plate. The vibration effect allows the coffee beans to be pushed onto the spiral bevel to capture images and screen them. Although the aforementioned color sorting machine solves the problem that is caused by screening in the falling process, the creator finds that there are still many shortcomings in the above-mentioned color sorting machine in implementation. Loss, first of all, the vibrating plate is generally made of metal material. Therefore, it is necessary to make other transparent materials on it, which is extremely complicated in process and difficult to commercialize. Secondly, the vibrating plate is in the process of vibration transmission. The coffee beans are transported by the vibration push. Therefore, the stacking of the coffee beans is actually easy to occur. In particular, when the spiral slope is narrow, the above-mentioned situation is more likely to occur, which affects the image quality.
承上，前述色選機僅是依靠顏色來達到辨識瑕疵豆的效果，但難以篩選出與正常豆顏色相近的瑕疵豆，例如，破裂豆、萎縮豆...等，因此，當瑕疵豆數量龐大時，便會造成其判斷精準度不佳；再者，由於螺旋斜面的區域較小，導致攝像裝置、剔除裝置...等均需受限於前述狹小區域，造成安裝與設置上的不便；最後，由於震動盤在持續震動的過程中，咖啡豆亦會隨著震動，因此，攝像裝置所擷取的影像往往不夠清晰，影響了後續判斷瑕疵豆的結果。綜上所述可知，目前用以篩選咖啡豆的裝置均不盡完善，因此，如何有效改善前述問題，即為本創作在此探討的一大課題。 According to the above, the color sorting machine only relies on color to achieve the effect of identifying cowpea, but it is difficult to screen out cowpeas that are similar in color to normal beans, for example, broken beans, shrinking beans, etc., therefore, when the number of cowpeas When it is huge, it will cause its judgment accuracy to be poor; in addition, because the area of the spiral bevel is small, the camera device, the rejection device, etc. are all limited by the aforementioned narrow area, which causes inconvenience in installation and installation. Finally, because the vibrating plate is in the process of continuous vibration, the coffee beans will also vibrate. Therefore, the images captured by the camera are often not clear enough, which affects the subsequent judgment of the cowpea. In summary, the devices used to screen coffee beans are not perfect. Therefore, how to effectively improve the above problems is a major topic for the creation of this creation.
有鑑於習知用以篩選咖啡豆的各種裝置，於實際使用上仍有諸多缺失，因此，創作人憑藉著多年來專業從事設計、加工及製造之豐富實務經驗，且秉持著精益求精的研究精神，在經過長久的努力研究與實驗後，終於研發出本創作之一種具有轉盤之咖啡豆篩選系統，期藉由本創作之問世，有效解決前述問題，令使用者擁有更佳的使用經驗。 In view of the various devices used to screen coffee beans, there are still many shortcomings in practical use. Therefore, the creators rely on the rich practical experience of design, processing and manufacturing for many years, and adhere to the spirit of research excellence. After a long period of hard work and experimentation, I finally developed a coffee bean screening system with a turntable. Through the creation of this creation, the above problems can be effectively solved, and the user has better experience in using.
本創作之一目的，係提供一種具有轉盤之咖啡豆篩選系統，以藉由該轉盤的獨立作業，能使其上的咖啡豆保持或趨近於靜止狀態，以 利於擷取影像作業，同時，透過人工智慧技術來篩選出不符標準的咖啡豆，該咖啡豆篩選系統包括一入料機構、一轉盤、至少一影像擷取裝置、一資訊處理裝置及至少一剔除機構，其中，該轉盤能接收該入料機構傳來的咖啡豆，且其能以自身軸心旋轉，以使複數顆咖啡豆彼此保持一間距，並形成串列態樣；該影像擷取裝置能擷取咖啡豆的一初始影像，又，該資訊處理裝置內至少設有一影像資料庫與一處理單元，該影像資料庫內儲存有複數個咖啡豆模型與參數，該處理單元內建有至少一學習演算模組，該學習演算模組能執行機器學習訓練功能或深度學習訓練功能，以能辨識出不符標準的咖啡豆，該處理單元能比對各該初始影像與各該咖啡豆模型與參數，並在判斷出不符標準的咖啡豆後，產生一排除訊號，以使該剔除機構能根據排除訊號，去除不符標準的咖啡豆。 One of the aims of the present invention is to provide a coffee bean screening system having a turntable, by which the coffee beans can be kept or approached to a standstill by independent operation of the turntable. Conducive to capturing image operations, and screening out non-standard coffee beans through artificial intelligence technology, the coffee bean screening system includes a feeding mechanism, a turntable, at least one image capturing device, an information processing device, and at least one rejection The mechanism, wherein the turntable can receive the coffee beans from the feeding mechanism, and can rotate with its own axis to keep a plurality of coffee beans at a distance from each other and form a tandem state; the image capturing device The image processing device can be configured with at least one image database and a processing unit, and the image database stores a plurality of coffee bean models and parameters, and the processing unit has at least a built-in a learning calculus module capable of performing a machine learning training function or a deep learning training function to identify non-conforming coffee beans, the processing unit capable of comparing each of the initial images with each of the coffee bean models a parameter, and after determining that the standard coffee beans do not conform to the standard, an exclusion signal is generated, so that the rejection mechanism can remove the non-conformity according to the exclusion signal Coffee beans.
為便 貴審查委員能對本創作目的、技術特徵及其功效，做更進一步之認識與瞭解，茲舉實施例配合圖式，詳細說明如下： For the sake of your review, you can make a further understanding and understanding of the purpose, technical features and efficacy of this creation. The examples are shown in the following diagram:
[習知] [知知]
無 no
[本創作] [This creation]
1‧‧‧咖啡豆篩選系統 1‧‧‧Coffee Bean Screening System
11‧‧‧入料機構 11‧‧‧Feeding agency
12‧‧‧轉盤 12‧‧‧ Turntable
13‧‧‧下方影像擷取裝置 13‧‧‧ below image capture device
14‧‧‧資訊處理裝置 14‧‧‧Information processing device
141‧‧‧影像資料庫 141‧‧‧Image database
143‧‧‧處理單元 143‧‧‧Processing unit
1431‧‧‧學習演算模組 1431‧‧‧Learning calculus module
15‧‧‧剔除機構 15‧‧ ‧ culling agency
16‧‧‧上方影像擷取裝置 16‧‧‧Top image capture device
17‧‧‧導正裝置 17‧‧‧ guiding device
18‧‧‧出料機構 18‧‧‧Distribution agency
C‧‧‧咖啡豆 C‧‧‧ coffee beans
第1圖係本創作之咖啡豆篩選系統；第2圖係本創作之處理單元執行訓練階段的流程圖；及第3題係本創作之處理單元執行運行預測階段的流程圖。 The first picture is the coffee bean screening system of the present creation; the second picture is the flow chart of the execution stage of the processing unit of the present creation; and the third problem is the flow chart of the execution prediction stage of the processing unit of the present creation.
近年來，隨著人工智慧機器學習領域的發展突飛猛進，透過機器學習(machine learning)與深度學習(Deep Learning)的模型訓練過程中，能夠將各種影像特徵，如：顏色、形狀、斑點...等特徵都同時納入判斷， 故能有效提升影像處理的精準度，因此，創作人特別將前述人工智慧技術結合至本創作內，畢竟，直到目前為止，尚未有實際產品結合了前述人工智慧技術以應用於咖啡豆篩選的領域中，合先陳明。 In recent years, with the rapid development of the field of artificial intelligence machine learning, various image features such as color, shape, and speckle can be obtained through the model training of machine learning and deep learning. And other features are included in the judgment. Therefore, the accuracy of image processing can be effectively improved. Therefore, the creator specifically combines the aforementioned artificial intelligence technology into the creation. After all, no practical product has been combined with the aforementioned artificial intelligence technology to be applied to the field of coffee bean screening. In the middle, the first Chen Ming.
本創作係一種具有轉盤之咖啡豆篩選系統，在一實施例中，請參閱第1圖所示，該咖啡豆篩選系統1至少由一入料機構11、一轉盤12、至少一影像擷取裝置(如：下方影像擷取裝置13或上方影像擷取裝置16)、一資訊處理裝置14及至少一剔除機構15，其中，該入料機構11能夠將複數顆咖啡豆C輸送至該轉盤12上，在此特別一提者，該入料機構11能夠為履帶、震動盤或其它輸送機構，只要其能夠將咖啡豆C輸送至轉盤12上，即為本創作所稱之入料機構11。 The present invention relates to a coffee bean screening system having a turntable. In an embodiment, as shown in FIG. 1 , the coffee bean screening system 1 comprises at least a feeding mechanism 11 , a turntable 12 , and at least one image capturing device. (eg, lower image capturing device 13 or upper image capturing device 16), an information processing device 14 and at least one rejecting mechanism 15, wherein the feeding mechanism 11 can transport a plurality of coffee beans C to the turntable 12 In particular, the feeding mechanism 11 can be a crawler belt, a vibrating plate or other conveying mechanism as long as it can transport the coffee beans C to the turntable 12, which is referred to as the feeding mechanism 11 of the present invention.
復請參閱第1圖所示，該轉盤12能以自身軸心進行旋轉，當複數顆咖啡豆C由該入料機構11被傳輸至該轉盤12時，往往有先後順序，而轉盤12是持續地轉動，因此，能使相鄰的咖啡豆C彼此保持一間距，又，由於該轉盤12僅是平穩地轉動，其並不會受到入料機構11的傳動影響，故，當咖啡豆C脫離入料機構11而被輸送至轉盤12上時，便會停留於轉盤12，在此情況之下，不會發生咖啡豆C相互堆疊之情事，令該轉盤12上的咖啡豆C能彼此分離並形成串列態樣，同時，令每一顆咖啡豆C都能保持靜止或接近於靜止狀態。在該實施例中，以該轉盤12呈透明狀(如：玻璃材質)為佳，以利後續提及的取像作業，但不以此為限，業者能根據實際需求，調整該轉盤12的材質。 Referring to FIG. 1 , the turntable 12 can be rotated by its own axis. When a plurality of coffee beans C are transferred from the feeding mechanism 11 to the turntable 12, there is often a sequence, and the turntable 12 is continuous. The ground rotates, so that the adjacent coffee beans C can be kept at a distance from each other, and since the turntable 12 is only smoothly rotated, it is not affected by the transmission of the feeding mechanism 11, so when the coffee beans C are separated When the feeding mechanism 11 is conveyed to the turntable 12, it stays on the turntable 12, in which case the coffee beans C are not stacked on each other, so that the coffee beans C on the turntable 12 can be separated from each other. Form a tandem pattern while allowing each coffee bean C to remain stationary or close to a stationary state. In this embodiment, the turntable 12 is preferably transparent (eg, glass material) to facilitate the subsequent image capturing operation, but not limited thereto, the operator can adjust the turntable 12 according to actual needs. Material.
復請參閱第1圖所示，該影像擷取裝置能擷取咖啡豆C的一初始影像，且當該影像擷取裝置位於該轉盤12的底面下方位置時，該影像 擷取裝置能作為下方影像擷取裝置13，又，由於該轉盤12為透明材質，因此，該下方影像擷取裝置13能經由該轉盤12而擷取到咖啡豆C的初始影像(後稱底面初始影像)，且因轉盤12上的咖啡豆C處於靜止狀態，故該底面初始影像能較清晰，有助於提高後續辨識效果。再者，該資訊處理裝置14能接收下方影像擷取裝置13傳來的底面初始影像，在該實施例中，該資訊處理裝置14內至少設有一影像資料庫141與一處理單元143，其中，該影像資料庫141內儲存有複數個咖啡豆模型與參數，前述咖啡豆模型與參數能夠為瑕疵豆的特徵，或是不同處理方法(如：日曬處理、水洗處理、密處理...等)的咖啡豆特徵，或是不同豆種(如：耶加雪菲(Yirgacheff)、藝妓(Geisha)、夏威夷可娜(Hawaii Kona)...等)，或是不同等級之咖啡豆的特徵，意即，本創作之咖啡豆篩選系統1除了能夠篩選出異物與瑕疵豆之外，甚至能夠對咖啡豆進行分類或分級。 Referring to FIG. 1 , the image capturing device can capture an initial image of the coffee bean C, and when the image capturing device is located below the bottom surface of the turntable 12 , the image The pick-up device can be used as the lower image capturing device 13 . Moreover, since the turntable 12 is made of a transparent material, the lower image capturing device 13 can capture the initial image of the coffee bean C via the turntable 12 (hereinafter referred to as the bottom surface). The initial image), and because the coffee bean C on the turntable 12 is in a stationary state, the initial image of the bottom surface can be clearer, which helps to improve the subsequent recognition effect. Furthermore, the information processing device 14 can receive the initial image of the bottom surface transmitted by the image capturing device 13 in the lower portion. In this embodiment, the information processing device 14 is provided with at least one image database 141 and a processing unit 143. The image database 141 stores a plurality of coffee bean models and parameters, and the coffee bean model and parameters can be characteristics of cowpea or different treatment methods (eg, solar treatment, water washing treatment, dense treatment, etc.) ) the characteristics of coffee beans, or different types of beans (such as: Yirgacheff, Geisha, Hawaii Kona, etc.), or the characteristics of different grades of coffee beans That is to say, the coffee bean screening system 1 of the present invention can even classify or classify coffee beans in addition to foreign bodies and cowpeas.
復請參閱第1圖所示，該處理單元143內建有至少一學習演算模組1431，該學習演算模組能執行機器學習(machine learning)訓練功能或深度學習(Deep Learning)訓練功能，以能辨識出不符標準的咖啡豆，在該實施例中，請參閱第2圖所示，首先，該處理單元143能執行訓練階段，其會先建立至少一人工智慧學習模型(如：監督與半監督式學習(Supervised and semi-supervised learning)演算法、強化學習(Reinforcement learning)演算法、卷積類神經網路(Convolutional Neural Network)演算法、隨機森林(Random forest)演算法...等)，並在該學習演算模組1431中輸入巨量資料，前述巨量資料能夠為咖啡豆影像資料與咖啡豆影像辨識參數，其中，咖啡豆影像資料能夠為整張圖片，或是圖片經由影像處理方法所產生的影像資 訊(如：色彩直方圖、輪廓、斑點、大小...等)。 Referring to FIG. 1 , the processing unit 143 has at least one learning calculation module 1431 capable of performing a machine learning training function or a deep learning training function. The non-standard coffee beans can be identified. In this embodiment, please refer to FIG. 2. First, the processing unit 143 can perform a training phase, which first establishes at least one artificial intelligence learning model (eg, supervision and half). Supervised and semi-supervised learning algorithms, Reinforcement learning algorithms, Convolutional Neural Network algorithms, Random forest algorithms, etc. And inputting a huge amount of data in the learning calculation module 1431, wherein the huge amount of data can be a coffee bean image data and a coffee bean image identification parameter, wherein the coffee bean image data can be an entire image, or the image is processed through the image. Image generated by the method (eg color histogram, outline, spot, size, etc.).
承上，復請參閱第1及2圖所示，該處理單元143會由學習演算模組測試影像辨識的正確率，以判斷影像辨識正確率是否足夠，當判斷結果為是，則將訓練完成的相關資訊(咖啡豆模型與參數)輸出並儲存至影像資料庫141中；當判斷結果為否，則使學習演算模組藉由調整影像辨識參數或其他方式而實現自我修正學習；如此，藉由重複上述步驟以完成訓練。又，請參閱第3圖所示，該處理單元143會執行運行預測階段，其能基於前述的學習演算模組輸入初始影像(在此為底面初始影像)與咖啡豆模型與參數，並比對底面初始影像與咖啡豆模型與參數，以進行預測性影像辨識，進而得到至少一個咖啡豆的識別資訊，以能判斷出不符標準的咖啡豆C，之後，該處理單元143會針對不符標準的咖啡豆C產生一排除訊號。在此聲明者，前述不符標準的咖啡豆C之意思，除了包含異物與瑕疵豆之外，若本創作之咖啡豆篩選系統1應用於分級上，亦能夠包含未達等級標準的咖啡豆，合先敘明。 As shown in FIGS. 1 and 2, the processing unit 143 tests the correctness of the image recognition by the learning calculation module to determine whether the image recognition accuracy is sufficient. When the judgment result is yes, the training is completed. The related information (coffee bean model and parameters) is output and stored in the image database 141; when the judgment result is no, the learning calculation module realizes self-correction learning by adjusting image recognition parameters or other means; Repeat the above steps to complete the training. Moreover, referring to FIG. 3, the processing unit 143 performs a running prediction phase, which can input an initial image (here, a bottom initial image) and a coffee bean model and parameters based on the aforementioned learning calculus module, and compare and compare The initial image of the bottom surface and the coffee bean model and parameters are used for predictive image recognition, thereby obtaining identification information of at least one coffee bean, so as to be able to determine the coffee beans C that do not conform to the standard, and then the processing unit 143 will target the coffee that does not conform to the standard. Bean C produces an exclusion signal. In this statement, the above-mentioned non-standard coffee bean C means that in addition to foreign matter and cowpea, if the coffee bean screening system 1 of the present invention is applied to the classification, it can also contain coffee beans that are not up to the standard. Explain first.
復請參閱第1圖所示，該剔除機構15能接收該資訊處理裝置14傳來的排除訊號，並去除不符標準的咖啡豆C，在該實施例中，該剔除機構15位於轉盤12上，且為噴嘴，其能吹出空氣，以將不符標準的咖啡豆C吹離該轉盤12，但不以此為限，在本創作之其它實施例中，該剔除機構15能為負壓吸引裝置，且能位於該轉盤12之頂面上方的區域，並能將不符標準的咖啡豆C吸出該轉盤12；或者，該剔除機構15能為推離裝置(如：推桿)，以能將不符標準的咖啡豆C推出該轉盤12；只要該剔除機構15能根據排除訊號，排除不符標準的咖啡豆C即可。 Referring to FIG. 1 , the culling mechanism 15 can receive the exclusion signal sent by the information processing device 14 and remove the non-conforming coffee bean C. In this embodiment, the culling mechanism 15 is located on the turntable 12 . And a nozzle capable of blowing air to blow the non-standard coffee beans C away from the turntable 12, but not limited thereto. In other embodiments of the present invention, the reject mechanism 15 can be a vacuum suction device. And can be located in the area above the top surface of the turntable 12, and can suck the non-standard coffee beans C out of the turntable 12; or, the reject mechanism 15 can be a push-off device (such as: putter), so that it will not meet the standard The coffee bean C is introduced into the turntable 12; as long as the reject mechanism 15 can exclude the coffee beans C that do not conform to the standard according to the exclusion signal.
另外，由於該底面初始影像僅為咖啡豆C的底面，因此，當瑕疵處位於咖啡豆的頂面時，則會無法被識別出來，故，為了能提高辨識咖啡豆的準確率，在該實施例中，復請參閱第1圖所示，該咖啡豆篩選系統1還能在該轉盤12的頂面上方位置設有影像擷取裝置，以作為一上方影像擷取裝置16，且該上方影像擷取裝置16亦能擷取咖啡豆C的初始影像(後稱頂面初始影像)，其中，該上方影像擷取裝置16能夠對應於下方影像擷取裝置13的位置(如第1圖所示)，但不以此為限，業者亦可根據實際需求，調整上方影像擷取裝置16的位置，使其不對應於下方影像擷取裝置13。又，該上方影像擷取裝置16會將該頂面初始影像傳送至資訊處理裝置14，使得該處理單元143能比對頂面初始影像與咖啡豆模型與參數，並在判斷出不符標準的咖啡豆C後，會產生對應的排除訊號，令剔除機構15能去除前述不符標準的咖啡豆C。 In addition, since the initial image of the bottom surface is only the bottom surface of the coffee bean C, when the crucible is located on the top surface of the coffee bean, it cannot be recognized, so in order to improve the accuracy of identifying the coffee bean, the implementation is performed. For example, as shown in FIG. 1 , the coffee bean screening system 1 can also be provided with an image capturing device at an upper position of the top surface of the turntable 12 as an upper image capturing device 16 , and the upper image The capturing device 16 can also capture the initial image of the coffee bean C (hereinafter referred to as the top surface initial image), wherein the upper image capturing device 16 can correspond to the position of the lower image capturing device 13 (as shown in FIG. 1). ), but not limited thereto, the operator can also adjust the position of the upper image capturing device 16 according to actual needs, so that it does not correspond to the lower image capturing device 13. Moreover, the upper image capturing device 16 transmits the top surface initial image to the information processing device 14, so that the processing unit 143 can compare the top surface initial image with the coffee bean model and parameters, and determines that the standard coffee is not satisfied. After the bean C, a corresponding exclusion signal is generated, so that the rejecting mechanism 15 can remove the aforementioned non-standard coffee bean C.
再者，創作人發現，當咖啡豆C由入料機構11輸送至轉盤12後，因其所具有的橢圓形外觀，往往會在轉盤12上滾動，因此，為了能使咖啡豆C盡量處於預定位置，方便下方影像擷取裝置13與上方影像擷取裝置16取得影像，復請參閱第1圖所示，該咖啡豆篩選系統1尚包括一導正裝置17，該導正裝置17會位在轉盤12上，並能將該入料機構11所輸送之咖啡豆C進行導正排列，令複數顆咖啡豆C能彼此分離並形成串列態樣，在該實施例中，該導正裝置17係為至少一擋板，其中，該擋板能呈一角度擺設，當咖啡豆C滾落至轉盤12並碰撞到該擋板後，其會受到擋板的阻擋而改變滾動方向，此時，加上該轉盤12的轉動，能夠使相鄰的咖啡豆C彼此分離並形成串列態樣，惟，在本創作之其它實施例中，該導正裝置17亦能為至少一滾輪， 滾輪同樣會呈一角度擺設，當咖啡豆C接觸到滾輪後，其能受到滾輪之推動及轉盤12的轉動，而同樣彼此分離並形成串列態樣。 Furthermore, the creator finds that when the coffee bean C is transported by the feeding mechanism 11 to the turntable 12, it tends to roll on the turntable 12 because of its elliptical appearance. Therefore, in order to make the coffee bean C as intended as possible The position is convenient for the lower image capturing device 13 and the upper image capturing device 16 to acquire an image. Referring to FIG. 1 , the coffee bean screening system 1 further includes a guiding device 17 , which is located at On the turntable 12, the coffee beans C conveyed by the feeding mechanism 11 can be aligned, so that the plurality of coffee beans C can be separated from each other and form a tandem state. In this embodiment, the guiding device 17 At least one baffle, wherein the baffle can be disposed at an angle. When the coffee bean C rolls down to the turntable 12 and collides with the baffle, it is blocked by the baffle to change the rolling direction. In addition, the rotation of the turntable 12 can separate the adjacent coffee beans C from each other and form a tandem pattern. However, in other embodiments of the present invention, the guiding device 17 can also be at least one roller. The rollers are also arranged at an angle. When the coffee beans C are in contact with the rollers, they can be pushed by the rollers and rotated by the turntable 12, and are also separated from each other and form a tandem pattern.
復請參閱第1圖所示，該咖啡豆篩選系統1還包括一出料機構18，其能接收來自該轉盤12傳來之符合標準的咖啡豆C，在該實施例中，該出料機構18係繪製成軌道搭配檔板的態樣，以使符合標準的咖啡豆C能受到檔板的阻擋而依序地進入軌道中，但不以此為限，只要該出料機構18能夠使轉盤12上的咖啡豆(符合標準)，被輸送至業者預期的區域，即屬於本創作所稱之出料機構18。 Referring to FIG. 1, the coffee bean screening system 1 further includes a discharge mechanism 18 capable of receiving the standard-compliant coffee beans C from the turntable 12. In this embodiment, the discharge mechanism The 18 series is drawn in the form of a track-matching baffle so that the standard-compliant coffee beans C can be sequentially entered into the track by the baffle, but not limited thereto, as long as the discharging mechanism 18 can make the turntable The coffee beans on the 12 (according to the standard) are delivered to the area expected by the operator, which is referred to as the discharge mechanism 18 referred to in this creation.
綜上所述可知，由於本創作之咖啡豆篩選系統1是採用轉盤12，且該轉盤12與入料機構11兩者是互不干擾的獨立裝置，因此，當咖啡豆C被輸送至轉盤12後，即能在轉盤12上保持於靜止或接近於靜止狀態，以方便下方影像擷取裝置13、上方影像擷取裝置16能清楚地取得咖啡豆的影像；又，該咖啡豆篩選系統1尚會以機器學習(machine learning)或深度學習(Deep Learning)來訓練資訊處理裝置14，以能辨識出咖啡豆的相關特徵，其中，機器學習最基礎的用法，是使用大量的數據和演算法來分析數據，以「訓練」機器從中學習，而深度學習則更進一步地透過大量多層數的類神經網路，使機器可由類神經網路自己學習找出重要的特徵資訊，但無論是機器學習或深度學習，在後續辨識咖啡豆C的成果上，均有效輔助人工辨識上的不足與效率，故能取得使用者的青睞；再者，本創作之轉盤12的空間區域會較習知技術的螺旋斜面更為寬廣，因此，業者能夠便於在前述空間區域中設置所需數量的下方或上方影像擷取裝置13、16及剔除機構15，且當前述裝置或機構故障或需檢修時，工作人員亦較有足夠空間作業。 In summary, since the coffee bean screening system 1 of the present invention uses the turntable 12, and the turntable 12 and the feeding mechanism 11 are independent devices that do not interfere with each other, when the coffee beans C are transported to the turntable 12 After that, it can be kept at a stationary state or close to a stationary state on the turntable 12, so that the lower image capturing device 13 and the upper image capturing device 16 can clearly obtain the image of the coffee bean; The information processing device 14 is trained by machine learning or deep learning to recognize the relevant characteristics of the coffee beans. The most basic usage of machine learning is to use a large amount of data and algorithms. Analyze the data, learn from the "training" machine, and deep learning further through a large number of multi-level neural networks, so that the machine can learn important information from the neural network itself, but whether it is machine learning Or deep learning, in the subsequent identification of the results of coffee beans C, are effective in assisting the deficiencies and efficiency of manual identification, so that users can be favored; The space area of the turntable 12 is wider than the spiral bevel of the prior art, so that the operator can conveniently set the required number of lower or upper image capturing devices 13, 16 and the culling mechanism 15 in the aforementioned spatial area, and When the above-mentioned device or mechanism is faulty or needs to be repaired, the staff also has more space to work.
按，以上所述，僅係本創作之較佳實施例，惟，本創作所主張之權利範圍，並不侷限於此，按凡熟悉該項技藝人士，依據本創作所揭露之技術內容，可輕易思及之等效變化，均應屬不脫離本創作之保護範疇。 According to the above description, it is only a preferred embodiment of the present invention, but the scope of the claims claimed by the present invention is not limited thereto, and according to those skilled in the art, according to the technical content disclosed in the present invention, Equivalent changes that are easily thought of should be in the protection of this creation.
Claims (12)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW108206223U TWM583944U (en) | 2019-05-17 | 2019-05-17 | A rotary-disk-based system for coffee bean sorting |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW108206223U TWM583944U (en) | 2019-05-17 | 2019-05-17 | A rotary-disk-based system for coffee bean sorting |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| TWM583944U true TWM583944U (en) | 2019-09-21 |
Family
ID=68620672
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW108206223U TWM583944U (en) | 2019-05-17 | 2019-05-17 | A rotary-disk-based system for coffee bean sorting |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWM583944U (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115227119A (en) * | 2022-08-01 | 2022-10-25 | 昆山亿政咖啡有限公司 | Intelligent coffee grinding method and system |
-
2019
- 2019-05-17 TW TW108206223U patent/TWM583944U/en unknown
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115227119A (en) * | 2022-08-01 | 2022-10-25 | 昆山亿政咖啡有限公司 | Intelligent coffee grinding method and system |
| CN115227119B (en) * | 2022-08-01 | 2024-02-06 | 昆山亿政咖啡有限公司 | Intelligent coffee grinding method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI714088B (en) | A rotary-disk-based system for coffee bean sorting | |
| US20080017557A1 (en) | System and Method for Sorting Larvae Cocoons | |
| US20110202169A1 (en) | Apparatus and method for sorting plant material | |
| KR102475930B1 (en) | Tobacco circumferential surface offline appearance quality inspection device and method | |
| CN112233107B (en) | Sunflower seed grade classification method based on image processing technology | |
| CN107392931A (en) | Bar tobacco board sorter and method | |
| CN109101976B (en) | Method for detecting surface defects of arc-extinguishing grid plate | |
| TWM583944U (en) | A rotary-disk-based system for coffee bean sorting | |
| CN211839115U (en) | Multi-index comprehensive spherical fruit and vegetable quality grading and sorting device | |
| JP2013505115A (en) | Process and equipment for sorting and boxing fruits | |
| CA2594281C (en) | System and method for sorting larvae cocoons | |
| US20200156119A1 (en) | Device and Method for Classifying Seeds | |
| US20090274811A1 (en) | Defect separation from dry pulp | |
| CN106991667A (en) | A kind of prawn integrality method of discrimination for building characteristics of image spectrum | |
| CN204523582U (en) | Coding type grain seed separator | |
| CN115254655A (en) | Multi-index passion fruit quality grading method based on machine vision | |
| CN107319583A (en) | Tea seed is peeled off separator | |
| KR102150313B1 (en) | Walnut automatic peeling apparatus | |
| CN208449937U (en) | The panorama potato defect detecting device combined is imaged based on actual situation | |
| CN112676179A (en) | Fruit quality detection grading equipment based on novel feeding and image capturing mode | |
| Kesiman et al. | Semi-automatic Ground Truth Image Construction for Coffee Bean Defects Classification Based on SNI 01-2907-2008 | |
| TWI796111B (en) | Screening machine for defective coffee beans and screening method thereof | |
| TWI541757B (en) | An agricultural product picking system | |
| CN114354628B (en) | Rhizome agricultural product defect detection method based on machine vision | |
| CN116558416A (en) | Flat object size detection device and method |