TW384451B - Counting system for shrimp and fish fry using mechanical visualization - Google Patents
Counting system for shrimp and fish fry using mechanical visualization Download PDFInfo
- Publication number
- TW384451B TW384451B TW85105254A TW85105254A TW384451B TW 384451 B TW384451 B TW 384451B TW 85105254 A TW85105254 A TW 85105254A TW 85105254 A TW85105254 A TW 85105254A TW 384451 B TW384451 B TW 384451B
- Authority
- TW
- Taiwan
- Prior art keywords
- fry
- vision
- machine
- counting
- image
- Prior art date
Links
Landscapes
- Farming Of Fish And Shellfish (AREA)
Abstract
Description
A7 B7 經濟部中央橾準局貝工消费合作社印輦 五、發明説明(1 ) ' 一種應用機器視覺之蝦魚苗計數糸統。 機器視覺技術是一門深具潛力,應用很廣的檢測技術, 目前在工業界上之應用相當普及。在國外視覺系統應用在 農業工程上典型的成功例子,包括農產品收稽自動化處理 5 工程(Slaughter and Harrell, 1987; Whittaker et al., 1987),以及農產品瑕班的檢測及分級(Rehkugler and Throop,1986*Berlage et al- ,1988iSarker and Wolfe» 1985:Wolfe and Swaminathan,1987) 〇 在國內櫬器視覺技術應用在農業工程的研究方面,有李 10 芳繁先生發表論文(1990)採用黑白視貴系統應用在檸樣幾 合形狀的悬測,並建立大小分級準則。另外,李芳繁先生 又發表「擰權頭色分级方法之建立」(1991),進一步使用 彩色視費系統來建立撺攆顔色分級方法。林達德先生與黃 俊誠先生發表「育苗箱種苗影像特性與背景雜訊之去除」 15 (1991),係使用黑白攝影視覺糸統應用在育苗箱上種苗缺 株之影像辨視分析。謝青霖先生等發表「數位影像處理在 蔬果大小選別之應用」(1992),研突藏果大小之影像測盪 技術。楊清富先生與李芳緊先生發表「應用機器視覺進行 番節顔色分級之研究」(1994),進一步應用彩色視覺糸統 20 建立一套蕃節分級機構。 台灌人工養殖事業相當發達,帶動蝦魚苗繁殖事業的蓬 勃發展,人工養殖魚螈成為國内食用魚蝦的主要來源。 锻魚苗經人工繁殖後,轉售放養至田間大養殖池,在轉 售過程須作計數價工作。現今锻魚苗的計數方法,目前均 (請先Η讀背面之注意事項再璘寫本霣) •裝.A7 B7 Printed by the Shellfish Consumer Cooperative of the Central Bureau of Quasi-Ministry of Economic Affairs. 5. Description of the invention (1) 'A shrimp fry counting system using machine vision. Machine vision technology is a promising technology with a wide range of applications. It is currently widely used in industry. Typical successful examples of the application of vision systems in agricultural engineering abroad, including the automated processing of agricultural products 5 (Slaughter and Harrell, 1987; Whittaker et al., 1987), and the detection and classification of agricultural product defect classes (Rehkugler and Throop , 1986 * Berlage et al-, 1988 iSarker and Wolfe »1985: Wolfe and Swaminathan, 1987) 〇 In the field of agricultural engineering research on the application of domestic vision technology, Mr. Li Fangfan published a paper (1990) The system is applied to the suspension measurement of lemon-like shape, and the size classification criterion is established. In addition, Mr. Li Fangfan also published the "Establishment of a Weighted Head Color Classification Method" (1991), which further used a color viewing fee system to establish a color classification method. Mr. Lin Dade and Mr. Huang Juncheng published "Removal of Seedling Box Image Characteristics and Background Noise" 15 (1991), which uses black-and-white photography visual system to analyze the image of seedlings lacking in the seedling box. Mr. Xie Qinglin and others published "Application of Digital Image Processing in Fruit and Vegetable Size Selection" (1992), and researched the technique of measuring the size of Tibetan fruits. Mr. Yang Qingfu and Mr. Li Fangjian published "Research on the Color Classification of Fanjie with Machine Vision" (1994), and further applied the color vision system to establish a Fanjie classification mechanism. The artificial irrigation business in Taiwan irrigation is quite developed, which has promoted the vigorous development of shrimp and fry breeding. Artificial farmed fish bream has become the main source of domestic fish and shrimp. Forged fish fry are artificially propagated and resold to a large breeding pond in the field. Counting is required during the resale process. At present, the counting methods of forged fish fry are currently (please read the precautions on the back before writing the script).
-C-C
線 本紙張又度適用中國國家揉準(CNS ) A4规格(210X297公釐) 83.3.10,000 I Α7 Β7 經濟部中央橾準扃負工消費合作杜印簟 五、發明说明(2 ) 採人工計數方法;先將全部搬魚苗收集集中分配到許多收 集商後,再抽取其中一筒來計數姻魚苗數目,乘以收集茼 數量,估算全部總锻魚苗數,在計數方式是採用一小碗盤 在收集茼内舀取蝦魚苗,限於目視能力,每次只能取出至 5 多五至六隻,然後一次接一次的計數累加,然而每個收集 商内的锻魚苗有一萬至二萬隻不等的數目,故計數工作相 當耗時,且計數結果完全取決於蓮氣(因僅抽樣一個收集 筒予以計數)及執行計數工作人貝之良知,故欠缺公平性 ,很難建立公平交易原則,且從事此行業的工作人貝呈現 10 人力老化,計數工作未能傳承恐有失傳之處。 由於鹏魚苗具有游動習性及重叠性,計數工作很難掌握 ,故至今始終沒有裉好的計數方法來取代人工計數。本發 明特應用機器視覺糸統來計數媳魚苗的影像處理方法,以 提供一種更為精確便利的鹏魚苗計數糸統。 15 本發明探討精確計数網魚苗數躉的影像處理方法,欲提 高锻魚苗計數的準確性,係採用二值(Binary)數擧形態 法(mathematical morphology)中浸触運算(erosion) 及膨脹蓮算(dilation)交替使用,如此可使重®的姬魚 苗予以分離,或修整鹏魚苗的外形使之平滑化,便於最後 20 的精確計數。 本發明重點在探討高可靠性、準確性及穩定性的影像處 理之可行性計數方法,Μ便配合锻魚苗分配機構,開發一 套能在現場使用的鹏魚苗計數視覺糸統。 本發明的主要目的在探討影像處理方法,配合二值形態 ---------¢-- (請先《讀背面之注$項再#寫本買>The threaded paper is again applicable to the Chinese National Standard (CNS) A4 (210X297 mm) 83.3.10,000 I Α7 Β7 Central Ministry of Economic Affairs, Standards, Work, and Consumption Cooperation Du Yin 簟 5. Description of the invention (2) Manual counting method ; After all the fish fry collections are collected and distributed to many collectors, one of them is taken to count the number of fry, and multiplied by the number of collected slugs to estimate the total number of forged fry. Shrimp fry are limited to visual ability. Only 5 to 5 to 6 fish can be taken out at a time, and the counts are added up one at a time. However, the forged fry in each collector ranges from 10,000 to 20,000. The counting work is quite time-consuming, and the counting result depends entirely on the lotus (because only one collection tube is sampled for counting) and the conscience of the person who performed the counting work, so it lacks fairness and it is difficult to establish a fair trading principle, and Working in this industry, Bei Bei presents 10 aging human resources, and there is a risk that the counting work could not be passed down. Because the penguin fry has swimming habits and overlap, it is difficult to master the counting work, so there is no good counting method to replace manual counting. The present invention applies a machine vision system to image the image processing method of sturgeon fry, in order to provide a more accurate and convenient penn fry count system. 15 The present invention discusses an image processing method for accurately counting net fry fry counts. To improve the accuracy of forging fry fry counts, erosion and expansion lotus arithmetic are used in the mathematical morphology of Binary. (Dilation) is used alternately, which can separate the heavy fry fry, or trim the shape of the peng fry to smooth it, so that it can be accurately counted in the last 20. The present invention focuses on the feasibility counting method of image processing with high reliability, accuracy and stability, and M cooperates with the forage fry distribution mechanism to develop a visual system for penguin fry counting that can be used on site. The main purpose of the present invention is to discuss the image processing method, in conjunction with the binary form --------- ¢-(please first read the note $ item on the back and then #write the book to buy>
線 本紙張又度逍用中钃國家梯準(CNS ) Α4规格(210X297公釐) 83.3.10,000 A7 B7 五、發明説明(3 ) ^ 學法及計數方法,作為精確計數鹏魚苗之用。 其主要技術闪容如T: (1) 自動計算二值化閥值之有效法。 (2) 二值形態學法中浸蝕與膨脹之交替使用法。 5 (3)有效的影像計數法(blobs counting),用來計數 锻魚苗。 並配合軟體技巧,结合較低成本的視覺糸統,Μ滿足線 上計數作業速度之要求。 兹將本發明的主要實施方法及»驗結果說明如下: 10 (1)膨像分割法(Image segment) p 影像分割是將主體與背景分雛,影像分剌法係Μ二值化& 方式處理。本發明以邊緣特微自動計算閥值法來自動計箄 閥值,使其具二值化效果。 Σ 2e(x,y) .f (x,y)The paper is also used in the China National Standard for Standards (CNS) A4 (210X297 mm) 83.3.10,000 A7 B7 V. Description of the invention (3) ^ The learning method and counting method are used to accurately count penguin fry. Its main technology flash capacity is T: (1) An effective method for automatically calculating a binary threshold. (2) Alternate use of erosion and swelling in binary morphology. 5 (3) Effective blobs counting, used to count forged fry. In combination with software skills and a low-cost visual system, M meets the requirements of online counting speed. The main implementation methods and test results of the present invention are explained as follows: 10 (1) Image segmentation p Image segmentation is to divide the subject from the background, and the image division method is M binarization & method deal with. In the present invention, the threshold value is automatically calculated by using the edge special micro automatic calculation threshold value method, so that it has a binary effect. Σ 2e (x, y) .f (x, y)
X Y 15 Τ〇 =-------- Σ Σθ (χ, y) Το:二值化閥值 其中 e(x,y) = | f(x+l,y)-f(x,y) ί + i f(x,y+l)-f(x,y) | (2)數學形態法 20 形態學集合最基本蓮算有膨脹(dilation)、收縮(erosion) 、閉蓮算(closing)、和開蓮算(opening)及细線(thinning)。 當膨脹和收縮成對使用,重複應用膨脹和收縮可 Μ 產生開運算和閉蓮算,而它們的特性是減少影像中比结構 元素小的綑節,而不畲影響到整體的幾何特性。 本纸張尺度適用中國國家揉準(CNS ) Α4規格(21〇Χ297公嫠) 83.3.10,000XY 15 Τ〇 = -------- Σ Σθ (χ, y) Το: Binarization threshold where e (x, y) = | f (x + l, y) -f (x, y ) ί + if (x, y + l) -f (x, y) | (2) Mathematical Morphology 20 The most basic morphological set is dilation, erosion, and closing , And opening and thinning. When expansion and contraction are used in pairs, repeated application of expansion and contraction can generate open and closed calculations, and their characteristics are to reduce the knots smaller than the structural elements in the image without affecting the overall geometric characteristics. This paper size applies to China National Standard (CNS) A4 specification (21〇 × 297 cm) 83.3.10,000
經濟部中夬搮準扃貝工消费合作社印裝 A3 (請先《讀背面之注意事項再填寫本頁) fPrinted by the China Ministry of Economic Affairs, Zhunbi Shellfish Consumer Cooperatives A3 (please read the precautions on the back before filling this page) f
A7 B7 經濟部中央橾率局員工消费合作社印製 五、發明说明(4 ) 開蓮算是先浸蝕後膨腺的蓮算,閉運算是先膨脹後浸蝕 的運算。使用圓結構做開蓮算可以平滑邊緣、打斷窄橋、 和消去小的島與形狀上尖銳的地方。使用閉蓮箄,則可以 平濟邊緣、箱合裂縫和細長的深坑、填補小的洞和邊緣上 5 的缺口。 本發明試驗的過程是將姻魚苗取像並予Μ二值化的圖像 後,當嘏魚苗個艘在夕卜形上被分離出一些毛邊细點,必須 修飾成平濟連貫的個雅,而單一锻魚苗被分割成不完整個 賊必須再度連成一體,另外當锻魚苗重•時須儘可能予以 10 分離,為克服這些問題,本發明係利用有效的二值形態學 的浸蝕運算與膨脹運算交替菫覆使用,Μ便後鑛計數工作 的進行。 (3)影像計數法 t本發明根據物體數fit及物體面積等資料,利用繡幅長法 15 (Run length)來處理,Μ計數多少個體並同時記錄每一個 體的面積,作為判斷魚苗幾隻重叠的依據。 〔資驗結果〕 本發明藉由機器視覺糸統的影像處理方法,精確計數蝦 魚苗的數目。圖說一為魚苗經過取像後的原始圖像,以下 20 將每一處理步驟及其结果逐項說明: (1)二值化之效果 麵說二係經計算閥值所得之二值化圖像,其结果相當理 想,從圖說二可Μ得知潛游的魚經二值化後變小,,故經二 值化後,浮游魚較大,潛游魚較小。 ---------_^-- (讀先Μ讀背面之注f項再埃窝本霣)A7 B7 Printed by the Consumer Cooperatives of the Central Bureau of Economic Affairs of the Ministry of Economic Affairs. 5. Description of the invention (4) Kailian is a lotus calculation that is eroded first and then swollen gland, and closed calculation is a calculation that swells first and then erodes. Using a round structure to open the lotus can smooth edges, interrupt narrow bridges, and eliminate small islands and sharp points in shape. With closed lotus roots, you can smooth edges, box cracks and slender pits, fill small holes and 5 gaps in the edges. The process of the test of the present invention is to take the image of the bristlefish fry and give it a binary image, and when the tuna fry are separated from each other in the shape of the burr, some fine points must be modified into a smooth, elegant, and A single forged fry is divided into incomplete thieves and must be reconnected as a whole. In addition, when the forged fry is heavy, it must be separated as much as possible. In order to overcome these problems, the present invention uses an effective binary morphological erosion operation and expansion. The calculation is used alternately, and the mine counting work is performed. (3) Image counting method t The present invention uses the embroidery length method 15 (Run length) to process according to the number of objects fit and area of the object. The number of individuals is counted and the area of each individual is recorded at the same time as the number of juveniles. Basis of overlap. [Inspection Results] The present invention accurately counts the number of shrimp and fry by using the image processing method of the machine vision system. The first picture is the original image of the fry after being taken. The following 20 describes each processing step and its results one by one: (1) The effect of binarization. The second is a binary image obtained by calculating the threshold. The result is quite satisfactory. It can be seen from the illustration that the snorkeling fish becomes smaller after binarization. Therefore, after binarization, the planktonic fish is larger and the snorkeling fish is smaller. ---------_ ^-(read first f read the note f on the back and then the nest)
订 線 本纸张尺度逋用中國國家標準(CNS ) Α4规格(210Χ297公釐) 83.3.10,000 經濟部t央標準局Λ工消费合作社印製 A7 B7 五、發明说明(5 ) - (2) 形態處理方法與結果 從圓說二中的二值圓像结果觀察可Μ得知,利用形態學 的浸蝕蓮算可將重叠魚苗予以分離。_說三是圖說二做浸 蝕運算的結果,從跚說三的结果顯示,一些重叠魚苗,經 5 浸触運算後可成功地予以分離。 此外,將浸蝕後的结果再行作膨脹運算,也就是倣開運 算,得到如圓說四的影像,其结果顯得比較理想。 但發現跚說四行開運算後,一些魚苗原先是完整個艘被 分雛出小支體,卻沒有被開蓮算中的膨膠過程連成一完整 10 個級,故若再經膨脹加收縮之閉運算可克服此問题,如圖 說五所示,而面積計算準則能照用的理由,係閉捶算是先 / 行膨脹再行收縮,故魚苗有將圆說四中被游離出小個艘再 度連成一體的效用,故經開閉運算可提高計數的準確性。 (3) 計數方法與結果’ 15 將形態處理後的影像,賴由面積計算或點數計算,Μ計 數出魚苗數目,其計弊準確率可達98%以上。且計數的 速度與人工計數比較,其計數速度高於人工計數為十數倍 甚至數十倍,因魚體的大小而略有不同,魚體愈小,其計 數能力愈高〇 20 Μ下再就本發明的機構部分,按_說明如下: 請閱第一、二、三圖所示,其係本發明之圓盤式機構示 意圖。其係在一平台1上以圓周方式平均分佈設置有若干 碗盤11,碗盤11受馬達12帶動可依軸心13旋轉,碗盤11上 方設有一具視覺機器2 ,藉由電腦3之控制,Μ輸出入控 (請先閏讀背面之注意事項再填寫本Ji)The paper size of the booklet is in accordance with the Chinese National Standard (CNS) A4 specification (210 × 297 mm) 83.3.10,000 Printed by the Central Bureau of Standards of the Ministry of Economic Affairs Λ Industrial Consumer Cooperative A7 B7 V. Description of the invention (5)-(2) Form processing Methods and results It can be seen from the observation of the binary circle image in Yuan Shuo Er that the overlapping fry can be separated by using the morphological erosion lotus calculation. _The third is the result of the etching operation of Figure Two. From the results of the first three, some overlapping fish fry can be successfully separated after the 5 immersion operation. In addition, the result of the etching is again subjected to the expansion operation, that is, the imitation operation is performed, and an image such as Yuan said is obtained, and the result looks ideal. However, it was found that after the four-line open operation, some fry were originally separated into small branches from the whole ship, but they were not connected to a complete 10 stages by the expansion process in the open lotus calculation. The closed operation can overcome this problem, as shown in Figure 5, and the reason why the area calculation criterion can be used is that the closed operation is first expanded / contracted. Therefore, the fry has been released into a small ship. Once again, the utility is integrated, so the accuracy of counting can be improved by the operation of opening and closing. (3) Counting method and result ’15 The morphologically processed image depends on area calculation or point calculation. M counts the number of fry, and the accuracy rate is over 98%. Compared with manual counting, the counting speed is ten or even dozens of times higher than the manual counting. It is slightly different due to the size of the fish. The smaller the fish, the higher its counting ability. Regarding the mechanism part of the present invention, the explanation is as follows: Please refer to the first, second, and third figures, which are schematic diagrams of the disc type mechanism of the present invention. It is arranged on a platform 1 with a plurality of bowls 11 evenly distributed in a circular manner. The bowls 11 are driven by a motor 12 to rotate around an axis 13. A visual machine 2 is arranged above the bowls 11 and controlled by a computer 3 , M output control (Please read the precautions on the back before filling in Ji)
本紙張尺度適用中國國家揉準(CNS ) Α4規格(210Χ297公釐) 83.3.10,000 Α7 Β7 五、發明説明(6 ) 制單元31可控制馬達運轉,進而帶動碗盤11旋轉,且Μ數 位影像處理單元32控制視覺機器2紈行影像取像處理;如 此周而復始,可逐次計數每一只碗盤11内的鹏魚苗數目。 而計數完成後的鹏魚苗則随碗盤11的肘節機構14,於傾斜 5 面15將網魚苗傾倒到下方的水槽内。 謫閲第四圓所示,其係本發明之另一種形態的輸送帶式 機構示意圈〇其係Μ—循環式輸送帶4上平均分佈設置若 千碗盤41,其他控制部分均與第一圓所示者雷同。随著輸 堵帶4的運轉,使輸送帶4上碗盤41內的锻魚苗,逐一地 10 被計數完成後,再自動傾倒於下方的水槽内。 綜上所述,本發明以櫬器視冕來處理锻魚苗的計數糸統 •係確實可供產業之利用,且符合新類性及進步性要件, 尚祈惠賜發明專利權為播。 1.1^^1 m nml nn <請先闻讀背面之注$項再填寫本頁)This paper size applies to China National Standard (CNS) A4 (210 × 297 mm) 83.3.10,000 Α7 Β7 V. Description of the invention (6) The control unit 31 can control the motor to drive the bowl 11 to rotate, and M digital image processing The unit 32 controls the vision machine 2 to perform image acquisition processing; in this way, the number of peng fry in each bowl 11 can be counted successively. After completion of the counting, the peng fry follow the elbow mechanism 14 of the bowl 11 and dump the net fry on the inclined surface 5 to the sink below. As shown in the fourth circle, it is a schematic circle of a conveyor belt mechanism of another form of the present invention. The M-circulating conveyor belt 4 is evenly distributed with thousands of dishes 41, and other control parts are the same as the first. Those shown in the circle are the same. With the operation of the congestion belt 4, the forged fry in the bowl 41 on the conveyor belt 4 are counted one by one, and then automatically poured into the lower water tank. To sum up, the present invention uses a sacrifice to treat the counting system of wrought fish fry. • It is indeed available to the industry and meets the new and progressive requirements. It is still hoped that the invention patent right will be broadcast. 1.1 ^^ 1 m nml nn < Please read the note on the back before filling in this page)
/IV- —線- 經濟部中央棣準局Λ工消费合作社印S. 83. 3.10,000 本紙張尺度遄用中國國家揉率(CNS ) Α4規格(210X297公釐} B7 五、發明説明(7 ) 圖式說明: 第1圖:本案發明之機構示意圖。 第2圖:本案發明之圓盤式機構俯視圖。 第3圖:本案發明之圓盤式機構正視圆。 第4圖:本案發明之另一·種機構示意圖。 圖號說明: 1 平台 11 碗盤 12 馬達 13 軸心 14 肘節機構 15 傾斜面 2 視覺機器 3 電腦 31 輸出入控制單元 32 數位影像處理單元 4 輸送帶 41 碗盤/ IV- —Line-Printed by the Central Bureau of Standards, Ministry of Economic Affairs, Λ Industrial Consumer Cooperative, S. 83. 3.10,000 This paper size uses China National Kneading Ratio (CNS) Α4 specification (210X297 mm) B7 V. Description of the invention (7 ) Description of the drawings: Figure 1: Schematic diagram of the mechanism of the invention of the present case. Figure 2: Top view of the disc mechanism of the invention of the present scheme. Figure 3: Front view of the disc mechanism of the invention of the present scheme. Figure 4: Another aspect of the invention of the present scheme A kind of mechanism diagram. Drawing number description: 1 platform 11 bowl 12 motor 13 axis 14 toggle mechanism 15 inclined surface 2 vision machine 3 computer 31 input / output control unit 32 digital image processing unit 4 conveyor belt 41 dish
(請先W讀背面之注f項再填寫本頁) 經濟部中央樣準局貝工消費合作社印裂 本紙張尺度適用中國國家標準(CNS ) A4規格(210X297公釐) 费) 83. 3.10,000 五、發明説明(牙 A7 B7 飄;_(Please read the note f on the back before filling out this page) The paper size printed by the Central Bureau of Standards, Ministry of Economic Affairs, Shellfish Consumer Cooperatives, is printed on the paper standard applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) fee 83. 3.10, 000 V. Description of the invention (tooth A7 B7 floating; _
園說一:魚苗的原始圖像 ----------•裝-- (請先《讀背面之注意事項再4寫本X)The garden said one: the original image of the fish fry ---------- • equipment-(please read the precautions on the back and then write 4 X)
-C.-C.
MBMB
V < 線 經濟部中央標隼局員工消费合作社印製 U 7 ◊ / _說二:二值化圓像 本紙張尺度適用中國國家標準(CNS ) A4规格 210X297公* ) α〇1 83.3.10,000 A7 B7 五、發明説明(7 ) 臟園 ^ 7 Ί 1 r^r / (請先閱讀背面之注意事項再填寫本筲)V < Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs U 7 ◊ / _ Say two: Binary circular image The paper size is applicable to the Chinese National Standard (CNS) A4 specification 210X297 male *) α〇1 83.3.10,000 A7 B7 V. Description of the invention (7) Dirty garden ^ 7 Ί 1 r ^ r / (Please read the notes on the back before filling in this note)
C 裝. 園說三:漤蚰運算的结果 % 經濟部中央標準局員工消費合作社印製 <二、夂 、-善、 7身· 線 圖說四:開運算的结果 本紙伕尺度適用中國國家標準(CNS ) A4规格(210 X 297公釐) iA) 83. 3.10,000 384451 A7 B7 五、發明説明(I。)C. Park said three: the result of the calculation of the% printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs < 2, 夂,-善, 7 body. (CNS) A4 specification (210 X 297 mm) iA) 83. 3.10,000 384451 A7 B7 V. Description of the invention (I.)
經濟部中央標準局員工消費合作杜印繁Consumption Cooperation among Staff of Central Bureau of Standards, Ministry of Economic Affairs
本紙張尺度適用中國國家樣準(CNS ) A4規格(210X297公釐) 813.10,000This paper size applies to China National Standard (CNS) A4 (210X297 mm) 813.10,000
Claims (1)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW85105254A TW384451B (en) | 1996-05-02 | 1996-05-02 | Counting system for shrimp and fish fry using mechanical visualization |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW85105254A TW384451B (en) | 1996-05-02 | 1996-05-02 | Counting system for shrimp and fish fry using mechanical visualization |
Publications (1)
Publication Number | Publication Date |
---|---|
TW384451B true TW384451B (en) | 2000-03-11 |
Family
ID=21625229
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW85105254A TW384451B (en) | 1996-05-02 | 1996-05-02 | Counting system for shrimp and fish fry using mechanical visualization |
Country Status (1)
Country | Link |
---|---|
TW (1) | TW384451B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108765448A (en) * | 2018-05-28 | 2018-11-06 | 青岛大学 | A kind of shrimp seedling analysis of accounts method based on improvement TV-L1 models |
-
1996
- 1996-05-02 TW TW85105254A patent/TW384451B/en not_active IP Right Cessation
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108765448A (en) * | 2018-05-28 | 2018-11-06 | 青岛大学 | A kind of shrimp seedling analysis of accounts method based on improvement TV-L1 models |
CN108765448B (en) * | 2018-05-28 | 2022-01-07 | 青岛大学 | Shrimp larvae counting analysis method based on improved TV-L1 model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104992430B (en) | Full automatic three-dimensional liver segmentation method based on convolutional neural networks | |
CN107230204B (en) | A kind of method and device for extracting the lobe of the lung from chest CT image | |
Corbit et al. | Fractal dimension as a quantitative measure of complexity in plant development | |
EP3708069A1 (en) | Estimating and predicting tooth wear using intra-oral 3d scans | |
CN106570871B (en) | Fuzzy C-mean algorithm carotid ultrasound image Internal-media thickness measurement method and system | |
Duan et al. | An automatic counting system for transparent pelagic fish eggs based on computer vision | |
Dhandra et al. | Analysis of abnormality in endoscopic images using combined HSI color space and watershed segmentation | |
Raitoharju et al. | Benchmark database for fine-grained image classification of benthic macroinvertebrates | |
CN101727537A (en) | Computer determining method of mammary gland CR image based on double visual angle information fusion | |
Esmaeili et al. | Scanning Electron Microscopy of the scale morphology in Cyprinid fish, Rutilus frisii kutum Kamenskii, 1901 (Actinopterygii: Cyprinidae) | |
Nayak et al. | A cellular automata based optimal edge detection technique using twenty-five neighborhood model | |
CN112233107B (en) | Sunflower seed grade classification method based on image processing technology | |
Wang et al. | Separation and identification of touching kernels and dockage components in digital images | |
Zhang et al. | Kappa loss for skin lesion segmentation in fully convolutional network | |
CN105160346A (en) | Tongue coating greasyness identification method based on texture and distribution characteristics | |
Min et al. | MRI images enhancement and tumor segmentation for brain | |
CN108280833B (en) | Skeleton extraction method for plant root system bifurcation characteristics | |
Li et al. | Image segmentation based on improved unet | |
Ismael et al. | Evaluation of Transfer Learning with CNN to classify the Jaw Tumors | |
TW384451B (en) | Counting system for shrimp and fish fry using mechanical visualization | |
CN104504757A (en) | Blood vessel wall mesh division method based on restriction point set method | |
Xia et al. | Retinal vessel segmentation via a coarse-to-fine convolutional neural network | |
Nanayakkara et al. | Automatic breast boundary segmentation of mammograms | |
Dhandra et al. | Classification of abnormal endoscopic images using morphological watershed segmentation | |
CN114549523A (en) | Single-step depth network-based automatic detection method for multiple raw teeth in center of curved surface layer graph |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GD4A | Issue of patent certificate for granted invention patent | ||
MM4A | Annulment or lapse of patent due to non-payment of fees |