基于Canny的改进图像边缘检测算法
|
摘要
边缘检测是图像处理领域中最重要的关键技术之一。针对经典边缘检测算法抗椒盐噪声性能较差及阈值选取适应性不强等问题,提出了一种基于Canny的算法架构,结合自适应中值滤波(Adaptive Median Filtering, AMF)、大津法(Otsu)以及最大熵法(Maximum Entropy Method,MEM)的改进图像边缘检测算法。该算法首先结合改进自适应中值滤波对图像降噪,从而在保留图像细节的同时较好地滤除了椒盐噪声干扰。而后利用基于Otsu和MEM提出的改进双阈值选取方法,获取自适应的高低阈值对图像边缘进行检测,边缘检测准确度可以达到96%以上。实验结果表明,本文算法在椒盐噪声干扰下针对背景复杂的图像有更好的边缘检测效果。
Edge detection is one of the most important technologies in image processing area. Aiming at the problems that classical edge detection algorithm cannot resist salt-pepper noise interference and the dual threshold selection adaptability is not strong enough, an improved image edge detection algorithm is proposed based on Canny algorithm, combined with adaptive median filtering(AMF), Otsu algorithm and maximum entropy method(MEM). The algorithm firstly combines the improved adaptive median filtering to denoise the image, so as to filter out the noise interference on the basis of preserving the image details. Then the improved dual threshold selection method combined with Otsu and MEM is used to obtain the high and low thresholds adaptively. The edge detection accuracy can reach 96% or more. The experimental results show that the proposed algorithm has a better edge detection effect on images with background under the interference of salt and pepper noise.
Edge detection is one of the most important technologies in image processing area. Aiming at the problems that classical edge detection algorithm cannot resist salt-pepper noise interference and the dual threshold selection adaptability is not strong enough, an improved image edge detection algorithm is proposed based on Canny algorithm, combined with adaptive median filtering(AMF), Otsu algorithm and maximum entropy method(MEM). The algorithm firstly combines the improved adaptive median filtering to denoise the image, so as to filter out the noise interference on the basis of preserving the image details. Then the improved dual threshold selection method combined with Otsu and MEM is used to obtain the high and low thresholds adaptively. The edge detection accuracy can reach 96% or more. The experimental results show that the proposed algorithm has a better edge detection effect on images with background under the interference of salt and pepper noise.
引文
[1]曾庆化,陈艳,王云舒,等.一种基于ORB的快速大视角图像匹配算法[J].控制与决策,2017, 32(12):2233-2239.ZENG Qing-hua, CHEN Yan, WANG Yun-shu, et al. A fast matching method for large viewpoint changes images based on ORB algorithm[J]. Control and Decision, 2017,32(12):2233-2239.
[2] Wang S, Sun S F, Guo Q, et al. Image edge detection based on rotating kernel transformation[C]. 7~(th)International Congress on Image and Signal Processing, 2014:397-402.
[3] Leordeanu M, Sukthankar R, Sminchisescu C. Generalized boundaries from multiple image interpretations[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(7):1312-1324.
[4] Lei T, Fan Y Y, Wang Y. Colour edge detection based on the fusion of hue component and principal component anal-ysis[J]. IET Image Processing, 2014, 8(1):44-55.
[5] Dollar P, Zitnick C L. Fast edge detection using structured forests[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(8):1558-1570.
[6] Lawend H O, Muad A M, Hussain A. Robust edge detection based on canny algorithm for noisy images[J].Journal of Theoretical and Applied Information Technology, 2017, 95(19):5104-5114.
[7] Liu L S, Liang F Q, Zheng J S, et al. Ship infrared image edge detection based on an improved adaptive Canny algorithm[J]. International Journal of Distributed Sensor Networks, 2018, 14(3):155014771876463.
[8] Krishna A S, Reddy B E, Pompapathi M. Nonlinear noise suppression edge detection scheme for noisy images[C].International Conference on Recent Advances and Innovations in Engineering, 2014:1-6.
[9] Farahanirad H, Shanbehzadeh J, Pedram M M, et al. A hybrid edge detection algorithm for salt-and-pepper noise[C].Proceedings of the International Multiconference of Engineering and Computer Science, 2011.
[10] Thirumavalavan S, Jayaraman S. An improved teachinglearning based robust edge detection algorithm for noisy images[J]. Journal of Advanced Research, 2016, 7(6):979-989.
[11]董昱,高云波,刘翔.改进的遗传算法在Canny算子阈值选取中的应用[J].兰州交通大学学报,2014,33(6):1-5.DONG Yu, GAO Yun-bo, LIU Xiang. Application of the improved genetic algorithm in threshold selection of Canny operator[J]. Journal of Lanzhou Jiaotong University,2014, 33(6):1-5.
[12] Fu W L, Johnston M, Zhang M J. Low-level feature extraction for edge detection using genetic programming[J].IEEE Transactions on Cybernetics, 2014, 44(8):1459-1472.
[13] Canny J. A computational approach to edge detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, 8(6):679-698.
[14] Hwang H, Haddad R A. Adaptive median filters:new algorithms and results[J]. IEEE Transactions on Image Processing, 1995, 4(4):499-502.
[15] Otsu N. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems, Man, and Cybernetics, 1979, 9(1):62-66.
[16] Wu N. The maximum entropy method[M]. Springer, 1997.
[2] Wang S, Sun S F, Guo Q, et al. Image edge detection based on rotating kernel transformation[C]. 7~(th)International Congress on Image and Signal Processing, 2014:397-402.
[3] Leordeanu M, Sukthankar R, Sminchisescu C. Generalized boundaries from multiple image interpretations[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(7):1312-1324.
[4] Lei T, Fan Y Y, Wang Y. Colour edge detection based on the fusion of hue component and principal component anal-ysis[J]. IET Image Processing, 2014, 8(1):44-55.
[5] Dollar P, Zitnick C L. Fast edge detection using structured forests[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(8):1558-1570.
[6] Lawend H O, Muad A M, Hussain A. Robust edge detection based on canny algorithm for noisy images[J].Journal of Theoretical and Applied Information Technology, 2017, 95(19):5104-5114.
[7] Liu L S, Liang F Q, Zheng J S, et al. Ship infrared image edge detection based on an improved adaptive Canny algorithm[J]. International Journal of Distributed Sensor Networks, 2018, 14(3):155014771876463.
[8] Krishna A S, Reddy B E, Pompapathi M. Nonlinear noise suppression edge detection scheme for noisy images[C].International Conference on Recent Advances and Innovations in Engineering, 2014:1-6.
[9] Farahanirad H, Shanbehzadeh J, Pedram M M, et al. A hybrid edge detection algorithm for salt-and-pepper noise[C].Proceedings of the International Multiconference of Engineering and Computer Science, 2011.
[10] Thirumavalavan S, Jayaraman S. An improved teachinglearning based robust edge detection algorithm for noisy images[J]. Journal of Advanced Research, 2016, 7(6):979-989.
[11]董昱,高云波,刘翔.改进的遗传算法在Canny算子阈值选取中的应用[J].兰州交通大学学报,2014,33(6):1-5.DONG Yu, GAO Yun-bo, LIU Xiang. Application of the improved genetic algorithm in threshold selection of Canny operator[J]. Journal of Lanzhou Jiaotong University,2014, 33(6):1-5.
[12] Fu W L, Johnston M, Zhang M J. Low-level feature extraction for edge detection using genetic programming[J].IEEE Transactions on Cybernetics, 2014, 44(8):1459-1472.
[13] Canny J. A computational approach to edge detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, 8(6):679-698.
[14] Hwang H, Haddad R A. Adaptive median filters:new algorithms and results[J]. IEEE Transactions on Image Processing, 1995, 4(4):499-502.
[15] Otsu N. A threshold selection method from gray-level histograms[J]. IEEE Transactions on Systems, Man, and Cybernetics, 1979, 9(1):62-66.
[16] Wu N. The maximum entropy method[M]. Springer, 1997.