Comparative Study of Multiple Image Edge Detection Operators Applied to Size Measurement of Passiflora edulia Sims

doi:10.3969/j.issn.1000-2561.2022.12.018

Abstract

Abstract:

Size measurement was a very important and indispensable link in the postpartum sorting stage of passion fruit. This paper described a variety of edge detection operators and image recognition steps based on edge detection algorithm. A fruit image acquisition platform was constructed by simulating fruit sorting site. In order to increase the difficulty, forty passion fruits with certain wrinkle on the surface were obtained by using fruit image acquisition platform and forty images shoot from the top were selected as the test material. Based on the Visual Studio 2017 integrated development environment, called the OpenCV 3.2.0 machine vision library, C++ programming language was used to detect the edge of forty passion fruit images by Canny, Sobel, Laplacian and Scharr operators. The size of passion fruit was measured by photoshop 7.0.1 software based on images as the approximation of real fruit size and according to the value to measure the detection accuracy of algorithm. The detection parameters of the four operators were automatically optimized, and the detection parameters of each operator, the detection accuracy of the four operators and the single fruit detection time were statistically analyzed by SPSS software. The test results showed that the measurement accuracy of passion fruit size using Canny operator under (70, 105, 3) parameters was the highest, the average time consumed of the single fruit measurement was 1.98 ms, it was the fastest of four operators, which can satisfy the actual demand for the intelligent sorting of passion fruit.

Key words: passion fruit (Passiflora edulia Sims), edge detection, operator

CLC Number:

LIU Xian, LIN Yingzhi. Comparative Study of Multiple Image Edge Detection Operators Applied to Size Measurement of Passiflora edulia Sims[J]. Chinese Journal of Tropical Crops, 2022, 43(12): 2554-2563.

Figures/Tables 11

References 37

[1]	杨锋, 黄永春, 何仁, 任仙娥, 陈锦姬. 百香果加工适性的测定及提高其出汁率的研究[J]. 食品研究与开发, 2008, 29(5): 82-86.
	YANG F, HUANG Y C, HE R, REN X E, CHEN J J. Determination of processing suitability of passion fruit and study on improving its juice yield[J]. Food Research and Development, 2008, 29(5): 82-86. (in Chinese)
[2]	BARBOSA P R, VALVASSORI S S, BORDIGNON C L, KAPPEL V D, MARTINS M R, GAVIOLI E C, QUEVEDO J, REGINATTO F H. The aqueous extracts of Passiflora alata and Passiflora edulis reduce anxiety-related behaviors without affecting memory process in rats[J]. Journal of Medicinal Food, 2008, 2(11): 282-288.
[3]	COLETA M, BATISTA M T, CAMPOS M G, CARVALHO R, COTRIM M D, DE LIMA T C M, DA CUNHA A P. Neuropharmacological evaluation of the putative anxiolytic effects of Passiflora edulis Sims, its sub-fractions and flavonoid constituents[J]. Phytotherapy Research, 2006, 12(20): 1067-1073.
[4]	SENA L M, ZUCOLOTTO S M, REGINATTO F H, SCHENKEL E P, DE LIMA T C M. Neuropharmacological activity of the pericarp of Passiflora edulis flavicarpa degener: putative involvement of c-glycosylflavonoids[J]. Experimental Biology & Medicine, 2009, 234(8): 967-975.
[5]	林开颜, 吴军辉. 基于计算机视觉的水果分级技术研究进展[J]. 信息化纵横, 2009(10): 1-4.
	LIN K Y, WU J H. Research progress of fruit grading technology based on computer vision[J]. Informatization Freely, 2009(10): 1-4. (in Chinese)
[6]	任永新, 单忠德, 张静, 战丽. 计算机视觉技术在水果品质检测中的研究进展[J]. 中国农业科技导报, 2012, 14(1): 98-103.
	REN Y X, SHAN Z D, ZHANG J, ZHAN L. Research progress of computer vision technology in fruit quality detection[J]. China Agricultural Science and Technology Guide, 2012, 14(1): 98-103. (in Chinese)
[7]	张书彦, 张文毅, 余山山, 付宇超. 图像处理技术在信息农业中的应用现状及发展趋势[J]. 江苏农业科学, 2017, 45(22): 9-13.
	ZHANG S Y, ZHANG W Y, YU S S, FU Y C. Application status and development trend of image processing technology in information agriculture[J]. Jiangsu Agricultural Sciences, 2017, 45(22): 9-13. (in Chinese)
[8]	邓泓, 杨滢婷, 刘志超, 彭莹琼. 基于实时数据的智能农业图像处理系统的实现[J]. 数字技术与应用, 2019, 37(11): 160-161.
	DENG H, YANG Y T, LIU Z C, PENG Y Q. Implementation of intelligent agricultural image processing system based on real-time data[J]. Digital Technology and Application, 2019, 37(11): 160-161. (in Chinese)
[9]	姚立健, 边起, 雷良育, 赵大旭. 基于BP神经网络的水果分级研究[J]. 浙江农业学报, 2012, 24(5): 926-930.
	YAO L J, BIAN Q, LEI L Y, ZHAO D Q. Research on fruit grading based on BP neural network[J]. Acta Agriculturae Zhejiangensis, 2012, 24(5): 926-930. (in Chinese)
[10]	周林妹. 数字图像边缘检测算法及其在农产品加工中的应用[J]. 食品与机械, 2009, 25(3): 139-153.
	ZHOU L M. Digital image edge detection algorithm and its application in agricultural product processing[J]. Food and Machinery, 2009, 25(3): 139-153. (in Chinese)
[11]	黄凤. 边缘检测算子在武夷岩茶图像处理中的比较[J]. 廊坊师范学院学报(自然科学版), 2013, 13(4): 34-37.
	HUANG F. Comparison of edge detection operators in Wuyi Rock Tea image processing[J]. Journal of Langfang Normal University (Natural Science Edition), 2013, 13(4): 34-37. (in Chinese)
[12]	刘现, 蔡淑芳, 黄语燕. 基于Canny算子的福橘图像边缘检测研究[J]. 东南园艺, 2018(6): 19-22.
	LIU X, CAI S F, HUANG Y Y. Research on edge detection of Kumquat image based on Canny operator[J]. Southeast Horticulture, 2018(6): 19-22. (in Chinese)
[13]	朱培逸, 刘红晴. 基于数学形态学图像分割算法在水果分级中的应用[J]. 科学技术与工程, 2013, 13(34): 10375-10380.
	ZHU P Y, LIU H Q. Application of image segmentation algorithm based on mathematical morphology in fruit grading[J]. Science Technology and Engineering, 2013, 13(34): 10375-10380. (in Chinese)
[14]	刘新庭, 匡迎春, 陈熵, 欧明文, 曾文辉. 基于最小外接圆直径的苹果分级研究[J]. 中国农学通报, 2015, 31(20): 239-243.
	LIU X T, KUANG Y C, CHEN S, OU M W, ZENG W H. Study on apple grading based on minimum circumscribed circle diameter[J]. Bulletin of Chinese Agronomy, 2015, 31(20): 239-243. (in Chinese)
[15]	崔巍, 高珏, 朱培逸, 徐金星, 邓婉. 基于机器视觉的金桔大小分级研究[J]. 常熟理工学院学报(自然科学), 2012, 26(4): 87-90.
	CUI W, GAO J, ZHU P Y, XU J X, DENG W. Research on size classification of Kumquat based on machine vision[J]. Journal of Changshu Institute of Technology (Natural Science), 2012, 26(4): 87-90. (in Chinese)
[16]	黄辰, 费继友. 基于图像特征融合的苹果在线分级方法[J]. 农业工程学报, 2017, 33(1):285-291.
	HUANG C, FEI J Y. Apple online grading method based on image feature fusion[J]. Journal of Agricultural Engineering, 2017, 33(1): 285-291. (in Chinese)
[17]	SNEHA A, SANYAM A. A new approach for image steganography using edge detection method[J]. International Journal of Innovative Research in Computer and Communication Engineering, 2013, 1(3): 626-629.
[18]	RASHMI, MUKESH K, ROHINI S. Algorithm and technique on various edge detection: a survey[J]. Signal & Image Processing, 2013, 4(3): 65-75.
[19]	JAMAL R. A novel automatic object detection in images[J]. Current Research in Computer Science, 2013, 1(1): 1-4.
[20]	LEEMANS V, DESTAIN M F. A real-time grading method of apples based on features extracted from defects[J]. Journal of Food Engineering, 2004(61): 83-89.
[21]	DEVRIM U, BERNARD G, OLIVIER K, VINCENT L, MARIE-FRANCE D, OLIVIER D. Automatic grading of bicolored apples by multispectral machine vision[J]. Computer and Electronics in Agriculture, 2011(75): 204-212.
[22]	HASSANKHANI R, NAVID H. Potato sorting based on size and color in machine vision system[J]. Journal of Agricultural Science, 2012, 4(5): 235-244.
[23]	ABDOLLAH G, FARID B, HOSSEIN B. Design, development and evaluation of an online potato sorting system using machine vision[J]. International Journal of Agriculture and Crop Science, 2013, 6(7): 396-402.
[24]	CANNY J. A computational approach to edge detection[J]. IEEE Trans Pattern Analysis and Machine Intelligence, 1986(12): 679-697.
[25]	林雯. 边缘检测算子分析与比较[J]. 湖南工业职业技术学院学报, 2012, 12(5): 6-25.
	LIN W. Analysis and comparison of edge detection operators[J]. Journal of Hunan Polytechnic College, 2012, 12(5): 6-25. (in Chinese)
[26]	周逢道, 马晨, 潘自超. 一种基于Sobel算子的星敏感器星图预处理方法[J]. 导航定位与授时, 2021, 8(4): 98-106.
	ZHOU F D, MA C, PAN Z C. A star map preprocessing method of star sensor based on Sobel operator[J]. Navigation, Positioning and Timing, 2021, 8(4): 98-106. (in Chinese)
[27]	冯伍, 张俊兰, 苗秋瑾. 几种典型边缘检测算子的评估[J]. 电子设计工程, 2011, 19(4): 131-133.
	FENG W, ZHANG J L, MIAO Q J. Evaluation of several typical edge detection operators[J]. Electronic Design Engineering, 2011, 19(4): 131-133. (in Chinese)
[28]	王文豪, 姜明新, 赵文东. 基于Canny算子改进的边缘检测算法[J]. 中国科技论文, 2017, 12(8): 910-915.
	WANG W H, JIANG M X, ZHAO W D. Improved edge detection algorithm based on Canny operator[J]. Chinese Scientific Papers, 2017, 12(8): 910-915. (in Chinese)
[29]	李静, 陈桂芬, 丁小奇. 基于改进Canny算法的图像边缘检测方法研究[J]. 计算机仿真, 2021, 38(4): 371-375.
	LI J, CHEN G F, DING X Q. Research on image edge detection method based on improved Canny algorithm[J]. Computer Simulation, 2021, 38(4): 371-375. (in Chinese)
[30]	毛星云, 冷雪飞. OpenCV3编程入门[M]. 北京: 电子工业出版社, 2015.
	MAO X Y, LENG X F. Introduction to OpenCV3 programming[M]. Beijing: Electronic Industry Press, 2015. (in Chinese)
[31]	欧温暖. 几种常见边缘检测算子的分析比较[J]. 现代计算机(专业版), 2010(5): 75-77.
	OU W N. Analysis and comparison of several common edge detection operators[J]. Modern Computer (Professional Edition), 2010(5): 75-77. (in Chinese)
[32]	彭士有, 吴洁, 彭俊. 拉普拉斯边缘检测算法的改进及其在探地雷达中的应用[J]. 现代雷达, 2020, 42(8): 41-45.
	PENG S Y, WU J, PENG J. Improvement of laplacian edge detection algorithm and its application on GPR[J]. Modern radar, 2020, 42(8): 41-45. (in Chinese)
[33]	杜伟, 何毅斌, 吴林慧, 汪强. 融合改进形态学和LOG算子的齿轮边缘检测[J]. 武汉工程大学学报, 2021, 43(6): 675-680.
	DU W, HE Y B, WU L H, WANG Q. Gear edge detection based on improved morphology and LOG operator[J]. Journal of Wuhan University of Engineering, 2021, 43(6): 675-680. (in Chinese)
[34]	乔闹生, 邹北骥, 邓磊, 曾友兵, 邹剑臣. 一种基于图像融合的含噪图像边缘检测方法[J]. 光电子激光, 2012, 23(11): 2215-2219.
	QIAO N S, ZOU B J, DENG L, ZENG Y B, ZOU J C. An edge detection method of noisy image based on image fusion[J]. Photoelectron Laser, 2012, 23(11): 2215-2219. (in Chinese)
[35]	张泽泓, 杨成忠, 李广亮. 交通标志边缘检测技术研究[J]. 杭州电子科技大学学报, 2014, 34(1): 83-86.
	ZHANG Z H, YANG C Z, LI G L. Research on edge detection technology of traffic signs[J]. Journal of Hangzhou University of Electronic Science and Technology, 2014, 34(1): 83-86. (in Chinese)
[36]	BRADSKI G, KAEHLER A. 学习OpenCV[M]. 北京: 清华大学出版社, 2009.
	BRADSKI G, KAEHLER A. Learning OpenCV[M]. Beijing: Tsinghua University Press, 2009. (in Chinese)
[37]	武松, 潘发明. SPSS统计分析大全[M]. 北京: 清华大学出版社, 2014.
	WU S, PAN F M. SPSS statistical analysis[M]. Beijing: Tsinghua University Press, 2014. (in Chinese)

算子 Operator	优点 Advantage	缺点 Disadvantage
Sobel	对噪声具有平滑抑制的作用^[26]	得到的边缘较粗且可能出现伪边缘^[26]
Prewitt	受噪声影响较小，适用于灰度渐变和低噪声图像^[27]	定位精度低^[27]
Roberts	比较适用于应用在水平和垂直边缘的检测，在对水平和垂直边缘定位时能够较准确地定位出位置^[24]	抗噪音能力弱，对噪音较敏感^[24]
Canny	抗干扰性强，计算复杂度低^[28]	人为设置高低阈值导致算法的适应性低，图像边缘信息丢失，产生伪边缘^[29]
Scharr	突变有较强的响应，速度快^[30]	阈值无法很好地分离边缘候选点中边缘点与非边缘点^[30]
Laplacian	定位精度高^[31]	不具有sobel和prewitt等边缘检测算子的图像平滑功能，对噪声的响应敏感,会误将噪声作为边缘^[32]
Log	边界定位精度高，抗干扰能力强，连续性好^[33]	对噪声敏感^[33]

算子 Operator	优点 Advantage	缺点 Disadvantage
Sobel	对噪声具有平滑抑制的作用^[26]	得到的边缘较粗且可能出现伪边缘^[26]
Prewitt	受噪声影响较小，适用于灰度渐变和低噪声图像^[27]	定位精度低^[27]
Roberts	比较适用于应用在水平和垂直边缘的检测，在对水平和垂直边缘定位时能够较准确地定位出位置^[24]	抗噪音能力弱，对噪音较敏感^[24]
Canny	抗干扰性强，计算复杂度低^[28]	人为设置高低阈值导致算法的适应性低，图像边缘信息丢失，产生伪边缘^[29]
Scharr	突变有较强的响应，速度快^[30]	阈值无法很好地分离边缘候选点中边缘点与非边缘点^[30]
Laplacian	定位精度高^[31]	不具有sobel和prewitt等边缘检测算子的图像平滑功能，对噪声的响应敏感,会误将噪声作为边缘^[32]
Log	边界定位精度高，抗干扰能力强，连续性好^[33]	对噪声敏感^[33]

算子 Operator	调优参数 Tuning parameters	步长 Step	范围 Range
Canny	Threshold1	5	5~100
	Threshold2	0.5 *threshold1	1.0 * threshold1- 5.0 * threshold1
	ApertureSize	2	3~7
Sobel	ksize	2	1~7
Sobel	delta	1	0~10
Laplacian	ksize	2	1~7
Laplacian	delta	1	0~10
Scharr	delta	1	0~10

算子 Operator	调优参数 Tuning parameters	步长 Step	范围 Range
Canny	Threshold1	5	5~100
	Threshold2	0.5 *threshold1	1.0 * threshold1- 5.0 * threshold1
	ApertureSize	2	3~7
Sobel	ksize	2	1~7
Sobel	delta	1	0~10
Laplacian	ksize	2	1~7
Laplacian	delta	1	0~10
Scharr	delta	1	0~10

算子 Operator	灰度图像 Grayscale image	二值图像 Binary image
Canny	NULL	NULL
Sobel
Laplacian
Scharr