食品專業(yè)畢業(yè)設(shè)計外文翻譯--利用計算機視覺和機器人進行食品處理和包裝

1、<p><b>　　本科畢業(yè)設(shè)計</b></p><p><b>　　外文文獻及譯文</b></p><p>　　文獻、資料題目：Food Handling Using Computer </p><p><b>　　文獻、資料來源：</b></p><p>　　文

2、獻、資料發(fā)表（出版）日期：2008.12</p><p><b>　　院 （部）： </b></p><p><b>　　專 業(yè)： </b></p><p><b>　　班 級： </b></p><p><b>　　姓 名： </b&

3、gt;</p><p>　　學 號： </p><p><b>　　指導教師： </b></p><p>　　翻譯日期：2009.5.15</p><p><b>　　外文文獻：</b></p><p>　　Food Handling and Packagin

4、g using Computer vision and Robot</p><p><b>　　Abstract</b></p><p>　　Even though the use of robot vision system in manufacturing sectors is now a commonplace, however, the technology

5、embodied in these devices is poorly matched to industrial needs of food processors. In particular, food processing imposes special demands upon machinery. For instance the vision sensor must be programmed to detect the p

6、osition of single and isolated object as well as overlapping or occluding objects. Special grippers have to be designed for handling of food articles such that they h</p><p>　　1. Introduction</p><

7、p>　　From the incoming down to the packaging lines, locating, recognizing and handling food objects are very important in food processing industry. These tasks are performed routinely in food industry mainly for qualit

8、y evaluation and product classification. Such tasks are very laboriously demanding and tend to rely heavily on role of the human operator [1]. Hands of workers using raw materials of animal origin can heavily be contam

9、inated with faecal and other micro-pathogenic organisms [2]. The s</p><p>　　Robots have successfully been applied in a wide range of food industries primarily dealing with well-defined processes and product

10、s not only because they are relatively clean and hygienic, also because of their flexibility, ruggedness and repeatability. This trend will continue to grow with the increasing scrutiny and regulatory enforcements such a

11、s and Hazard Analysis and Critical Control Points (HACCP) together with companies that are looking for ways to decrease or eliminate worker exposure t</p><p>　　Firstly the food products, despite of the same

12、type, differ in size, shape and other physical variables. This imposes special demands for machinery to handle them,requiring multiple sensory, manipulation and environmental capabilities beyond those available in robots

13、 designed to automate manufacturing tasks. Secondly the success of applying robots for food handlers, hinges upon the success of detecting, locating, recognizing and handling severely overlapping and occluding cases of s

14、imilar food o</p><p>　　2. Materials and Methods </p><p>　　2.1 Sample Preparation </p><p>　　The chosen food for this study is a locally produced beef-burger. It possesses all the imp

15、ortant characteristics which are unique to food products, such that they are very fragile and easily deformed. The average size of the beef-burgers is 8.5 mm in thickness and 46.1 mm radius and 69.3 gm in weight. Surface

16、 images of test samples were acquired using 8-bit robot vision system with uniform white background. The white background provides excellent contrast between the burger and the background. T</p><p>　　2.2 Rob

17、ot vision </p><p>　　The robot vision systems used in this study is the Adept Cobra 600 4-DOF articulated scara robot, manufactured by Adept Tech., USA and equipped with Adept Vision Interface, MV-5 Adept con

18、troller and TM1001 CCD monochrome camera manufactured by Pulnix Inc., Canada. The camera was mounted onto link 2 of robot arm and illuminated using the warm white deluxe (WWX) fluorescent lighting. The camera is fitted w

19、ith a C-mount adapter to permit the use of Tamron f/25.5 8-mm lens. The TM1001 camera is conn</p><p>　　2.3 Image Processing </p><p>　　The objective of image processing in robot vision applicatio

20、ns is mainly to extract meaningful and accurate information from the images, endowing the robots with more sophisticated position control capabilities through the use of vision feedback. The use of a simple geometric me

21、thod such as introducing specially designed cues into the image scene will not work in this application since the burger images are generally complex, difficult to model and partially or extensively occluded depending on

22、</p><p>　　In order to accurately translate burger positions to robot movements, the former geometric features must firstly be extracted and secondly matched to the robot's workspace. In this application

23、one of the useful features which uniquely characterize the pose of a burger in arbitrary locations is its centroid. This geometric descriptor is applicable since the shape of a burger is approximately circular. Furthermo

24、re this feature preserves variance to translation, rotation and scaling. Before computi</p><p>　　Edge detection operation is carried out to detect the contour of the connected and isolated components, thereb

25、y, effectively transforming the original data into a form suitable for further processing. The edge results of Figure 2 computed using well-known Sobel and Robert operators [5] are shown in Figures 3(a),(b),(c)&(d).

26、From these figures it can be seen that the edges determined by these operators comprised of many false edges, discontinuities and spurious spots resulting from uneven and irr</p><p>　　A more sophisticated me

27、thod is needed in order to obtain acceptable results. The method used to solve these problems was based on Canny edge detection operator [6]. Interested readers are referred to this publication for detailed mathematical

28、 explanation of this relatively new edge detector. Here only the important principles are presented in order to facilitate discussion on robot vision applications on food handling. Canny method for edge detection is prin

29、cipally based on some general ideas.</p><p>　　Firstly Canny was the first to demonstrate that convolving an image with a symmetric 2-D Gaussian filter and then, secondly, differentiating in the direction of

30、the gradient form the edge magnitude image. The presence of edges in the original image gives rise to ridges in gradient magnitude image. The objective is to detect the true edge in the right place. This can be done usin

31、g method known as non-maximal suppression technique. Essentially this method works by tracking along the top of the rid</p><p>　　Comparing Figure 3 and Figure 4, it can be seen clearly that the edges determi

32、ned by Canny's operator are less corrupted compared to edges detected either by Sobel or Robert operator. The burger edges are more complete in Figure 4 whereas in Figure 3 they are only partially visible and more o

33、bscured. Furthermore the retention of major detail by the Canny operator is very evident. The presence of overlapping and partially occluding burgers are visually recognizable. Canny operator therefore has</p>&l

34、t;p>　　2.4 Centroid Detection Algorithm </p><p>　　Once the edges of the burgers have been detected, the next step in image analysis is to retrieve and extract the geometric feature which uniquely defines

35、the shape of a burger. One important criterion of this type of shape analysis and retrieval problems is that the method must be invariant to translation, scaling and translation of images or objects. The use of Hough tra

36、nsform seems to be adequate since this method achieves translation, scaling and rotation invariance by converting a global de</p><p>　　However the original Hough transform works well if analytic equations of

37、 object borderlines are known and invariant. In the present context these conditions are very difficult to be fulfilled because the shape of the burger is not a perfect circle. This imperfection is mainly due to non-rigi

38、d properties of the burgers, causing them to be easily deformed when pressed or come in contact with any rigid surface such as the conveyor belts. A straightforward application of Hough transform will yield a</p>

39、<p>　　Figure 5 CCD Curve of Circle</p><p>　　The basic idea of this technique can be explained using illustration in Figure 5. It shows a point Q lying on the contour of a circle which is characterized

40、by a centroid C, and a radius R. The angle between the point Q and the centroid is given by ? Tracing a burger contour can be considered as circling around its centroid. The tracing path in clockwise or anticlockwise dir

41、ection from fixed starting point represents a shape contour uniquely. In other words a contour point sequence corresponds </p><p>　　Next the number of instances the circumference points of that circle are al

42、so edges in the binarised image are determined. If the number of pixels are greater than a threshold, it implies that the centroid pixel being considered is the centroid of the burger. </p><p>　　Practically

43、 when the above algorithm is applied to burger images, the total number of matches would never be the maximum even for a correct centroid. This is because of inevitable noise, irregular light reflection and burger surfac

44、e shadows. Thus the threshold value for the number of matches has to be fixed below the maximum value 36.</p><p>　　To determine the correct value of the threshold the algorithm was applied to sequence of 19

45、burger images. The criteria for correctly identifying which species of the burger is most likely to be lifted are that, that burger should be minimally overlapped or maximally exposed. As seen in Figure 2 the burger that

46、 lies on the top of the heap as well as on the side of the main pile could fulfill the above criterion, and hence, contributed also as the picked-and-place species. The algorithm is applie</p><p>　　By follow

47、ing this criterion the robot will be led to pick and place only those species, thereby reducing the likelihood of damaging the overlapped specie. Figure 7 shows the number of matches of each burger centroid in a sequence

48、 of images using Eq. 2. </p><p>　　3. Experimental Tests and Results </p><p>　　The methods and procedures described in the previous sections were experimented using sequence of burger images. &

49、lt;/p><p>　　The objective of this experiment is to sort the burger individually by pick-and-place operations. In so doing the robot must first examine the present of burgers in the heap, and second, detect whic

50、h species of burger that was most likely to be lifted-up. Prior to experiment, the camera was calibrated for a given mounting position,enabling robot pose with respect to the position and orientation of the burger be acc

51、urately mapped. </p><p>　　Figures 8(b)-(f) show the sequence of centroids of minimally overlapped burgers revealed using modified Hough transform, starting with detection of the 1st burger and ending with d

52、etection of the 7th burger respectively. Only the first seven centroid locations were shown here even though the locations of a total of 19 burgers were successfully located. Each centroid location was fed into a control

53、ler which kinematically positioned and orientated the robot's end-effecter in 3D space. In each dete</p><p>　　It can therefore be concluded that the proposed method works well for detecting minimally ove

54、rlapping burgers which is important in ensuring a correct pick-and-place sequence of the robot. However, one drawback of this technique is that it is a very computationally intensive method, requiring approximately 3-4 s

55、econds for every result. A time consuming yet accurate position detection algorithm may limit its applications in food industry. Hence, a special hardware for fast position detection is n</p><p>　　Moreover a

56、 specially designed end-effecter is needed in meeting the need for robotic handling of beef-burgers. Clearly the use of conventional grippers is not suitable since they do not address the task of handling non-rigid mater

57、ials and they can increase contamination problems of beef-burgers. In order to solve these problems a novel non-contact end effecter employing pneumatic levitation technique [9] is now being investigated in our laborat

58、ory.</p><p>　　4. Conclusion </p><p>　　Technological development in robotics has the potential to minimize the contamination risk in food handling and packaging. The successful application of thi

59、s relatively new technology in food industry, however, requires compliant with several processing parameters, namely recognition of overlapping and touching objects. In this paper we have implemented a relatively simple

60、but effective recognition of overlapping and touching objects for use in robot positioning and guidance. By using global ima</p><p>　　The algorithm resulted form this study in modified Hough transform. This

61、algorithm was tested for detection of beef-burgers, and it was discovered that, the system is particularly robust, converging to the desired pose corresponding to a minimally overlapped burger or maximally exposed burger

62、 from initial pose over the entire work space. The algorithm has a very good accuracy in detecting the food objects with more than 10% overlapping or occlusion. Further extensions of this work include improv</p>&

63、lt;p>　　5. Acknowledgements </p><p>　　This work is supported by Malaysia Intensified Research in Priority Areas grant IRPA 6012602.</p><p>　　6. References </p><p>　　[1] KHODABANDE

64、HLOO, K., CLARKE, P.T. 1993. Robotics in meat. Fish and Poultry Processing. Chapman and Hall, London. </p><p>　　[2] DE-WIT, J.C. 1995. The importance of hand hygiene in contamination of foods. Antonie Van L

65、eeuwenhuek 51, 523-527. </p><p>　　[3] TRICKETT, J. 1992. Food hygiene for food handlers. Macmillan, Basingstoke, UK. </p><p>　　[4] LEGG, B. 1993. Hi-tech agricultural engineering - a contradicti

66、on in terms of the way forward. Mechanical Incorporated Engineer, August, 86-90. </p><p>　　[5] GONZALEZ, R.C. and WOODS, R.E. 2002. Digital image processing. Prentice Hall, USA. </p><p>　　[6] CA

67、NNY, J.F. 1986. A computational approach to edge detection. IEEE Transactions Pattern Recognition and Machine Intelligence, 8(6), 679-698. </p><p>　　[7] ILLINGWORTH, J. and KITTLER, J. 1988. A survey of the

68、Hough transform. Computer Vis., Graphics and Image Process., 44, 87-116. </p><p>　　[8] WANG, Z., CHI, Z. and FENG, D. 2003. Shape based leaf retrieval. IEE Proceedings Vis. Image Signal Process., 150(1), 34

69、-42. </p><p>　　[9] ERZINCANLI, F. SHARP, J. and ERHAL, S. 1997. Design and Operational Considerations of a Non-contact Robotic Handling System for Non-rigid Materials. Int. J. Mech. Tools Manufact., 38, 353-

70、361.</p><p><b>　　中文譯文：</b></p><p>　　利用計算機視覺和機器人進行食品處理和包裝</p><p><b>　　1 摘要</b></p><p>　　雖然在制造業(yè)使用機器人視覺系統(tǒng)已經(jīng)相當普及,然而這一技術(shù)在機械設(shè)備的應(yīng)用很難符合食品加工的工業(yè)需要.特別是,食品加

71、工對機械有特殊要求. 例如視覺傳感器必須程序化的檢測單一孤立的對象以及重疊或遮擋物體的位置。處理食品必須專門設(shè)計夾子以便于與食物有最小的接觸引起最小的損害。在這個項目為期一年的研究中，視覺導向系統(tǒng)正在開發(fā)以滿足本目標。該系統(tǒng)集成了修改后的版本的Hough變換算法作為主要的識別引擎。這一方法和程序在商業(yè)牛肉漢堡包上進行了測試。</p><p><b>　　2 介紹</b></p>

72、<p>　　從入境到包裝生產(chǎn)線，定位、識別、處理食物對食品加工業(yè)非常重要。這些任務(wù)一般在食品加工工業(yè)中主要用于質(zhì)量評價和產(chǎn)品分類。這些任務(wù)有極高的要求，而且往往在很大程度上依賴于人的操作。使用動物原材料的手工工人經(jīng)常被糞便和其他微生物病原微生物污染。一項特里克特的研究在食物中毒和食品加工衛(wèi)生標準在之間已顯示出強有力的聯(lián)系。高度自動化的食物處理和包裝用機械手是消除由人工處理引起的食品的微生物的質(zhì)量問題的最有效的手段。</

73、p><p>　　機器人已被成功廣泛的應(yīng)用在各種食品工業(yè)，主要處理規(guī)定的產(chǎn)品流程，這不僅是因為它們是相對清潔和衛(wèi)生，也因為他們的靈活性、耐用性和可重復性。這一趨勢將繼續(xù)增加審查、監(jiān)管</p><p>　　和執(zhí)法，如危害分析與關(guān)鍵控制點（HACCP）并結(jié)合。公司正在尋找新途徑來減少或消除工人的操作和重復動作以及所處的惡劣環(huán)境。但是在食品工業(yè)使用機器人也帶來了些許問題與挑戰(zhàn)。</p>

74、<p>　　首先是糧食產(chǎn)品，盡管是同一類型，但不同大小、形狀或其他物理量對機械處理都有相應(yīng)的特殊要求。這需要多種感官，操縱和環(huán)境的能力之外，還可以在機器人設(shè)計自動化生產(chǎn)任務(wù)。 其次成功的運用機器人對于食品操作員來說，取決于成功探測、定位、識別和處理嚴重重疊和遮擋情況下類似的食物對象。第三，食品對象往往是微妙的，而且通常覆蓋著滑或粘性物質(zhì)，使高速處理他們非常具有挑戰(zhàn)性。現(xiàn)有的那些機械裝置，如真空吸塵器和夾持鉗不適用，因為他們有可

75、能對所加工的食品造成損傷和擦傷。因此，為了解決這些問題需要進一步研究。本文解決了這樣一些的問題，例如用視覺傳感器來控制機器人，企圖以模擬人類眼睛的方式，來控制手臂的運動。</p><p><b>　　3 材料與方法</b></p><p><b>　　3.1樣品制備</b></p><p>　　這項研究所選擇的食物，是一個

76、本地生產(chǎn)的牛肉漢堡。它具有所有食物典型的特征，例如非常脆弱，易變形。一般大小的牛肉漢堡包為厚度8.5毫米、半徑46.1毫米和重量69.3克。測試樣本的表面圖像由具有統(tǒng)一的白色背景的8位機器人視覺系統(tǒng)操作獲得。白色背景與漢堡形成了鮮明的對比。調(diào)整所選擇的樣品，使圖像的強度直方圖大約圍繞在中間的全面范圍。調(diào)節(jié)焦距被選中，讓單一的以及多個樣品在圖像幀的范圍內(nèi)。</p><p><b>　　3.2 機器人視覺&

77、lt;/b></p><p>　　在此研究中使用的機器人視覺系統(tǒng)是Adept Cobra 600具有4個自由度的多關(guān)節(jié)機器人。該機器人視由美國的Adept Tech研制,運用獨特的視覺接口 、MV - 5控制器和由加拿大的Pulnix公司生產(chǎn)的TM1001CCD黑白照相機。相機裝在機械臂的鉸鏈2上，用豪華熒光燈照明。攝像機安裝帶有C -片匣允許使用騰龍f/25.5 8毫米的鏡頭。TM1001相機通過一個12

78、引腳廣瀨型數(shù)碼相機連接器（中廣公司，日本）連接到該視頻卡。機器人視覺系統(tǒng)由Adept’s AIMS v4.0和編程圖書館進行操作，運行在1.7 GHz和255 MB內(nèi)存的奔騰4電腦。圖1顯示了機器人視覺系統(tǒng)的安裝設(shè)置。</p><p>　　圖1 試驗設(shè)備的安裝</p><p><b>　　3.3 圖像處理</b></p><p>　　在機器人

79、視覺應(yīng)用中圖像處理的目標主要是從圖片中提取有意義和準確的信息，賦予機器人更先進的位置控制能力通過視覺反饋。采用一個簡單的幾何方法，如利用專門設(shè)計的線索采集現(xiàn)場的圖像將不會在這方面的工作，因為漢堡應(yīng)用圖像通常是復雜的，困難的模型。圖2顯示了典型的牛肉漢堡圖片。</p><p>　　圖2典型的牛肉漢堡圖片。</p><p>　　為了準確地獲得漢堡位置以使機器人動作，必須提取以前的幾何特征然后讓

80、其與機器人的工作范圍相匹配。在此應(yīng)用中的一個有用的特征是它的質(zhì)心。這是漢堡在任意位置的唯一的特征。這種幾何描述使用于描述近似圓形的漢堡。此外此特征保留平移、旋轉(zhuǎn)和縮放的差異。在對實際漢堡圖像質(zhì)心進行計算機處理之前，需要對每個圖像進行預處理。邊緣檢測操作進行了檢測輪廓的連接和分離的組成部分,從而，有效地轉(zhuǎn)化成原始數(shù)據(jù)形式適合作進一步處理。圖2計算眾所周知索貝爾和羅伯特,運營商[ 5 ]顯示在圖3（ a ）項，（b）項,（c）和（d）。這些

81、數(shù)字可以看出，所確定的邊緣這些運營商包括許多虛假邊緣，不連續(xù)性和雜散點造成的不平衡和不規(guī)則表面的漢堡包，非均勻輕反射和陰影。這些缺點不可以接受的應(yīng)用本文介紹。</p><p>　　圖3邊緣檢測結(jié)果（一）索貝爾邊緣低閾值，（二）Sobel邊緣高閾值，（三）羅伯特邊緣低閾值和（四）羅伯特邊緣較高的門檻</p><p>　　為了獲得可以接受的結(jié)果需要一個更先進的方法。所采用的方法被用來解決基于C

82、anny邊緣檢測運營商的問題。感興趣的讀者對這一出版物提出了相對較新的邊緣檢測的數(shù)學解釋。在討論應(yīng)用機器人視覺處理食物時在這里只提出了重要的原則。精明的邊緣檢測方法主要是基于一些普通的想法。</p><p>　　首先坎尼首先表明，convolving圖像對稱二維高斯過濾器，然后，第二，差異化的方向梯度形成規(guī)模的邊緣圖像。那個在場的邊緣在原始圖像產(chǎn)生脊坡度規(guī)模形象。其目標是發(fā)現(xiàn)真正的優(yōu)勢在恰當?shù)牡胤?。這是使用非最大

83、抑制技術(shù)的方法可以做到的。從本質(zhì)上講，這種方法通過跟蹤上方的山脊，只保留那些點上方的峰值，同時壓制所有其他國家。那個跟蹤過程展品滯后控制的兩個重要的參數(shù)。它們是低閾值Tlow ，上閾值Thigh。如果邊緣反應(yīng)首先是Thigh，然后肯定這個像素構(gòu)成的優(yōu)勢，因此保留。像素不到Thigh，但大于Tlow被視為弱邊緣。為了彌補所有停產(chǎn)的邊緣，以及消除不實最后跟蹤做。如果它們連接強大的優(yōu)勢保留弱邊。由于這些行動是一個形象薄線的邊緣點與改進的邊緣信

84、噪比，盡管這種方法降低了效果噪音，不過，整體素質(zhì)的邊緣主要是依賴優(yōu)化選擇的標準偏差Ⅰ。它規(guī)定了高斯遮罩坎尼的邊緣檢測。實驗的最佳值被設(shè)置為3。這相當于25 × 25內(nèi)核。此值固定給定的背景照明和圖像增益。改變?nèi)缯彰?，圖像的增益，背景顏色等這些外部因素，也將影響到最佳值的I 。圖4 （ a ）及（ b ）顯示結(jié)果Canny邊緣檢測與i設(shè)置為1和3 。</p><p>　　圖4使用Canny邊緣檢測的邊緣結(jié)

85、果（ a ）= 1 ，（ b ）= 3 </p><p>　　比較圖3和圖4可以看出明確指出，確定坎尼邊緣的算子與損壞邊緣檢測相比，無論是羅伯特索貝爾或經(jīng)營者。邊緣的漢堡包完成在圖4中，而在圖3中市部分可見的和模糊的。此外， 保留主要詳細的Canny算子是非常顯而易見的。存在重疊和部分阻斷漢堡視覺識別。因此，運營商有能力檢測漢堡形象的主要特征，幾何特點是取放物種是一個漢堡包可以準確地確定。該算法確定取放種給出了

86、下面一節(jié)。</p><p><b>　　3.4質(zhì)心探測算法</b></p><p>　　一旦邊緣的漢堡被檢測,圖像分析的下一步就是分析檢索和提取以定義形狀的漢堡包的獨特的幾何特征。這種類型的形狀分析和檢索問題的一個重要的標準是，該方法必須是不變的翻譯，擴大和翻譯圖片或?qū)ο?。使用Hough變換似乎是充分的，因為這種方法實現(xiàn)了翻譯，縮放和旋轉(zhuǎn)不變性轉(zhuǎn)換,檢測中存在的問題的

87、形象空間變成更容易的解決當?shù)貑栴}的峰值檢測參數(shù)空間。更重要的是，Hough變換允許分割重疊或半自動閉塞物體，對于處理漢堡包圖像這是至關(guān)重要的。</p><p>　　如果解析方程邊界的物體是已知和不變的，那原來的Hough變換工程就可以實現(xiàn)其任務(wù)。在目前情況下這些條件是很難實現(xiàn)，因為漢堡的形狀不是一個完美的圓。這個缺陷主要是由于漢堡非剛性性能的影響，使他們解除任何剛性表面時很容易變形如在傳送帶上。一個簡單的應(yīng)用

88、Hough變換將產(chǎn)生多重的積累票的參數(shù)空間，相應(yīng)的不同形狀和大小的物體。這可能導致許多假警報。此外，含糊不清的索引和不足，說明點功能可能會導致錯誤的解決方案，承認重疊或半封閉的物體。在這項工作中，采用最新的技術(shù)修改Hough變換在識別物體基于心輪廓距離（防治荒漠化公約）曲線的方法以解決一些問題。下圖給出了CCD方法。</p><p>　　圖5圓曲線的防治荒漠化公約這一技術(shù)的基本思想可圖5解釋說明。這表明點Q位于

89、由圓心C和半徑R決定的輪廓圓上。點Q和圓心的夾角是由S筆表現(xiàn)以此追查漢堡包輪廓，可視為盤旋在其圓心周圍。從固定的起點追查道路順時針或逆時針的方向，是一個獨特的形狀輪廓換言之等高點序列相當于一個形狀獨特的，如果出發(fā)點是固定。因此，對某一ç ，研究和S點Q的輪廓將準確地滿足下列標準，如果輪廓屬于一個完美的循環(huán)，即為Q = （ RCos? ， RSin? ）因為在這種情況下，一個完美的比賽是無法獲得如前所述的原因，因此，點Q是視為一

90、個點屬于邊緣漢堡包如果是范圍內(nèi)的最大和最小R值。</p><p>　?。?RminCosθ ，RminSinθ ）（ RmaxCosθ ，RmaxSinθ ）圖6顯示的漢堡的半徑范圍是Rmin <R< Rmax。這種方法的工作原理是首先對待所有邊緣像素的圖像造成binarized 從Canny邊緣檢測，因為可能centroids的物體。對于每一個質(zhì)心的位置，其次，防治荒漠化公約曲線追蹤使用均衡器。追

91、蹤S是各不相同在0 º360 º ，從而尋找所有像素這是這兩個之間的范圍內(nèi)輪廓。如果增加值為s的均衡器，二是保持1則最大的可能像素能夠滿足作為周長點為重心的360 。一般規(guī)模較小的上述某些價值觀限額提高計算時間的算法。對于每一個像素在binarised 形象正在考慮假定位中心的一個圓圈。</p><p><b>　　圖6漢堡半徑</b></p><p&

92、gt;　　下一步的一些情況下，腰圍點的循環(huán)還邊在binarised形象決心。如果像素數(shù)大于1閾值，這意味著目前的重心像素認為是漢堡的質(zhì)心。實際上，在上述算法應(yīng)用于以漢堡包圖像，總?cè)藬?shù)的比賽將永遠的最高得正確的圓心。這個正因為不可避免的噪音，不規(guī)則的光反射和漢堡表面陰影。因此，閾值比賽的數(shù)要低于固定最高值36。以確定正確的價值的閾。該算法適用于序列19漢堡包圖像。正確的標準，確定哪些物種的漢堡包是最有可能被取消的，這應(yīng)該是起碼的漢堡包重疊

93、或最大限度地暴露出來。正如圖2漢堡包的位于上方的堆積以及一側(cè)的主要能滿足上述標準，因此，挑選時相當重要的。那個算法是用于心位置最左右下角像素。因此多個漢堡包的符合標準的將被取消優(yōu)先從左上角至右下角。按照這一標準的機器人將領(lǐng)導挑選和地點只有那些物種，從而減少的可能性，損害了重疊。圖 7顯示了若干場檢測的每一個漢堡包心在一系列的圖像使用均衡器。</p><p><b>　　圖7閾值</b><

94、;/p><p>　　從圖可以得出結(jié)論，閾值的26制作了準確的檢測在挑選和地點種方面。</p><p>　　4 實驗測試和結(jié)果前面所描述的方法和程序用一系列的漢堡圖像來試驗驗證。</p><p>　　圖片（ a ）-（ g ）項的序列centroids 最小重疊漢堡檢測的上述算法。細圓表明位置刪除漢堡包的位置。這個實驗的目的是對單個漢堡進行挑揀分類。做這樣的機器人必須首

95、先研究目前的漢堡包的位置，第二發(fā)現(xiàn)該品種的漢堡包。之前的實驗，相機校準某一安裝位置，使機器人方面構(gòu)成中的地位和定位準確映射漢堡包。</p><p>　　圖片8（ b ） - （ F ）的顯示順序centroids 顯示使用重疊漢堡包 進行Hough轉(zhuǎn)換，開始檢測第一漢堡，最后檢測第七漢堡包。僅7圓心地點即使在這里顯示的地點共有19個位于漢堡。每個質(zhì)心位置被輸入一個控制器，其中動定位和面向機器人的最終效應(yīng)在三維空間

96、。在每個檢測輪的取放漢堡是手動。顯然從圖8的位置重疊或微最大限度地暴露漢堡包是準確地顯示在每一個選擇和放置循環(huán)。部分重疊或閉塞漢堡被發(fā)現(xiàn)。因此可以得出結(jié)論認為，擬議的行之有效的方法檢測微重疊漢堡這一點非常重要，以確保正確的選擇，。然而，這一技術(shù)的一個缺點是，它是一個計算非常密集的方法，大約需要3-4秒。一個費時尚未精確位置檢測算法，可以限制在食品中的應(yīng)用行業(yè)。因此，一個特殊的硬件快速立場檢測正在開發(fā)使用外地可編程門陣列（ FPGA ）芯

97、片。</p><p>　　另外一個特別設(shè)計的最終效應(yīng)是需要在滿足需要機器人處理的牛肉漢堡包。顯然使用常規(guī)夾子不合適的，因為它們是處理非剛性材料的，他們可以增加對牛肉漢堡包的污染。為了解決這些問題發(fā)明的新型非接觸式效應(yīng)氣動懸浮技術(shù)正在我們試驗室探究。</p><p><b>　　5結(jié)論</b></p><p>　　機器人技術(shù)的發(fā)展利用是為盡量減少

98、對食品處理和包裝潛在的污染，其已成功應(yīng)用這個相對較新的食品工業(yè)中，然而，需要符合幾個工藝參數(shù)，即承認重疊和觸摸物體。在本文中，我們已經(jīng)實施了相對簡單的但有效的重疊物體用于機器人定位和導引。通過利用全面的形象描述與先進圖像處理，對典型問題的步驟提取和匹配的幾何特征消除，從而能夠準確地定位機器人臂，即使存在嚴重的重疊情況。</p><p>　　由于該算法在本研究中修改了Hough變換，測試檢測的牛肉漢堡市健全的，融合

99、的理想構(gòu)成相應(yīng)的最低漢堡包或重疊最大限度地暴露漢堡包從最初的構(gòu)成比整個工作空間。該算法具有很好的準確檢測食品物體超過 10 ％的重疊或閉塞。進一步延長這一工作包括改進機器人最終效應(yīng)和運動學控制方案提供不間斷議案適用于需要動態(tài)跟蹤。 </p><p><b>　　6鳴謝</b></p><p>　　這項工作得到了馬來西亞加緊研究所的大力支持。7 參考資料</p&

100、gt;<p>　　[ 1 ] KHODABANDEHLOO 光克拉克制藥公司 1993年《機器人在肉類、魚和家禽加工中的應(yīng)用》查普曼和霍爾 倫敦 [ 2 ] DE-WIT J.C. 1995 《手部衛(wèi)生對食品污染的重要性》 安東尼范 51 523-527 。 [ 3 ]TRICKETT 美國 1992《食物從業(yè)員的食品衛(wèi)生》 麥克米倫 貝辛斯托克 英國[ 4 ] LEGG 1993《高科技農(nóng)業(yè)工程-