正余弦优化算法在多阈值图像分割中的应用
|
摘要
阈值的选取对图像分割后的效果有重要的影响,在传统的图像分割中存在分割结果单一,灵活度不强,以及容易陷入局部最优的问题。为了确定图像分割的最佳阈值,本文针对图像分割过程中涉及的阈值选取问题,提出一种基于正余弦优化算法(sine cosine algorithm,SCA)的多阈值图像分割方法。该算法以最大类间方差作为正余弦算法的适应度函数,通过正弦函数和余弦函数的变化来更新当前解在每一维度上的位置,候选解利用多个随机算子围绕最优解进行正余弦的波动来完成每一次的寻优过程,通过迭代计算更新最优解的位置,从而确定图像分割的最佳阈值。选取4幅标准测试图像进行实验,通过与粒子群优化算法进行峰值信噪比、结构相似法和寻优时间3方面的对比,结果表明:将正余弦优化算法应用在图像分割中可以获得更准确的分割阈值和更高的分割效率,具有很强的实用性。
Threshold selection has an important impact on the effect of image segmentation. In traditional image segmentation,there are some problems such as single segmentation result,weak flexibility and easy to fall into local optimum. In order to determine the optimal thresholds for image segmentation,a multilevel-thresholding image segmentation method based on sine cosine algorithm( SCA) is proposed in this paper. The method takes the between-class variance as the fitness function of the sine-cosine algorithm,updates the position of the current solution in each dimension by changing the sine function and cosine function. The candidate solution uses multiple random operators to carry out sine-cosine fluctuation around the optimal solution to complete each optimization process,and updates the optimum through iterative calculation. The optimal thresholds of image segmentation are determined by the location of the solution. Four standard test images are selected for experiment and compared with PSO algorithm in peak signal to noise ratio,structural similarity method,and optimization time. It shows the application of sine-cosine algorithm in image segmentation can obtain more accurate threshold and higher segmentation efficiency,which has a strong practicability.
Threshold selection has an important impact on the effect of image segmentation. In traditional image segmentation,there are some problems such as single segmentation result,weak flexibility and easy to fall into local optimum. In order to determine the optimal thresholds for image segmentation,a multilevel-thresholding image segmentation method based on sine cosine algorithm( SCA) is proposed in this paper. The method takes the between-class variance as the fitness function of the sine-cosine algorithm,updates the position of the current solution in each dimension by changing the sine function and cosine function. The candidate solution uses multiple random operators to carry out sine-cosine fluctuation around the optimal solution to complete each optimization process,and updates the optimum through iterative calculation. The optimal thresholds of image segmentation are determined by the location of the solution. Four standard test images are selected for experiment and compared with PSO algorithm in peak signal to noise ratio,structural similarity method,and optimization time. It shows the application of sine-cosine algorithm in image segmentation can obtain more accurate threshold and higher segmentation efficiency,which has a strong practicability.
引文
[1]李磊.图像分割研究现状概述[J].信息技术与信息化,2015,22(3):85-87.LI L. An overview of the research on image segmentation[J]. Information Technology and Informatization,2015,22(3):85-87.
[2]李玲,张旭,贾磊磊.火焰图像分割算法研究[J].科技视界,2015,5(33):105-106.LI L,ZHANG X,JIA LL. Research on flame image segmentation algorithm[J]. Science and Technology Horizon,2015,5(33):105-106.
[3]刘超,蔡文华,陆玲.图像阈值法分割综述[J].脑知识与技术,2015,11(1):140-142.LIU C,CAI W H,LU L. An overview of image segmentation threshold method[J]. Brain Knowledge and Technology,2015,11(1):140-142.
[4]申铉京,潘红,陈海鹏.基于一维Otsu的多阈值医学图像分割算法[J].吉林大学学报,2016,54(2):344-348.SHEN X J,PAN H,CHEN H P. Muti-threshold medical image segmentation algorithm based on one-dimensional otsu[J]. Journal of Jilin University,2016,54(2):344-348.
[5]程玉柱,蔡云飞.基于分数阶CV模型的木材缺陷图像分割算法[J].林业机械与木工设备,2018,46(4):44-47.CHENG Y Z,CAI Y F. Wood defect image segmentation algorithm based on fractional order CV model[J]. Forestry and Woodworking Equipment,2018,46(4):44-47.
[6]陈峥,石勇鹏,吉书鹏.一种改进的Otsu图像阈值分割算法[J].激光与红外,2012,42(5):584-588.Chen Z,Shi Y P,Ji S P. An improved threshold Segmentation for otsu images[J]. Laser and Infrared,2012,42(5):584-588.
[7]吴军,王龙龙.基于双鸟群混沌优化的otsu图像分割算法[J].微电子学与计算机,2018,35(12):119-124.WU J,WANG L L. Otsu image segmentation algorithm based on double Bird Group chaos optimization[J]. Microelectronics and Computer,2018,35(12):119-124.
[8]赵乃刚,邓景顺.粒子群优化算法综述[J].科技创新导报,2015,12(26):216-217.ZHAO N G,DENG J S. An overview of particle swarm optimization algorithm[J]. Scientific and Technological Innovation Guide,2015,12(26):216-217.
[9]李海楠,张学良.基于群体适应度方差的粒子群优化算法[J].计算机仿真,2007,24(5):158-161.LI H N,ZHANG X L. Particle swarm optimization algorithm based on variance of population fitness[J]. Computer Simulation,2007,24(5):158-161.
[10] MIRJALILI S. SCA:a sine cosine algorithm for solving optimization problems[J]. Knowledge-Based Systems,2016,96:120-133.
[11]刘勇,马良.转换参数非线性递减的正弦余弦算法[J].计算机工程与应用,2017,53(2):1-5.LIU Y,MA L. A sinusoidal cosine algorithm for nonlinear decrement of transfer parameters[J]. Computer Engineering and Application,2017,53(2):1-5.
[12]孙俊,宋佳,武小红.基于改进Otsu算法的生菜叶片图像分割方法[J].江苏大学学报,2018,39(2):179-184.SUN J,SONG J,WU X H. Segmentation of raw vegetable leaf image based on improved otsu algorithm[J]. Journal of Jiangsu University,2018,39(2):179-184.
[13]王裔,段会川. Otsu方法在多阈值图像分割中的应用[J].计算机工程与设计,2008,29(11):2844-2845.WANG Y,DUAN H C. Application of otsu method in multithreshold image segmentation[J]. Computer Engineering and Design,2008,29(11):2844-2845.
[14] OTSU N. A threshold selection method from gray-level histogram[J]. IEEE Transactions on System, Man, and Cybemetics,1979,9(1):62-66.
[15]何庆元,韩传久.基于粒子群算法的Otsu法图像阈值分割[J].桂林电子科技大学学报,2006,26(5):355-358.HE Q Y,HAN C J. Application of otsu method based on particle swarm optimization in image segmentation[J]. Journal of Guilin University of Electronic Science and Technology,2006,26(5):355-358.
[16]马文,周绍光,丁慧珍.基于遗传算法的Otsu法在图像分割中的应用[J].沿海企业与科技,2005,10(8):160-161.MA W,ZHOU S G,DING H Z. Application of otsu method based on genetic algorithm in image segmentation[J]. Coastal Enterprises and Science and Technology,2005,10(8):160-161.
[17] SHEIKH H R,SABIR M F,BOVIK A C. A statistical evaluation of recent full reference image quality assessment algorithms[J]. IEEE Transactions on Image Processing,2006,15(11):3441-3452.
[18] WANG Z,BOVIK A C,SHEIKH H R,et al. Image quality assessment:from error visibility to structural similarity[J]. IEEE transactions on image processing,2004,13(4):600-612.
[2]李玲,张旭,贾磊磊.火焰图像分割算法研究[J].科技视界,2015,5(33):105-106.LI L,ZHANG X,JIA LL. Research on flame image segmentation algorithm[J]. Science and Technology Horizon,2015,5(33):105-106.
[3]刘超,蔡文华,陆玲.图像阈值法分割综述[J].脑知识与技术,2015,11(1):140-142.LIU C,CAI W H,LU L. An overview of image segmentation threshold method[J]. Brain Knowledge and Technology,2015,11(1):140-142.
[4]申铉京,潘红,陈海鹏.基于一维Otsu的多阈值医学图像分割算法[J].吉林大学学报,2016,54(2):344-348.SHEN X J,PAN H,CHEN H P. Muti-threshold medical image segmentation algorithm based on one-dimensional otsu[J]. Journal of Jilin University,2016,54(2):344-348.
[5]程玉柱,蔡云飞.基于分数阶CV模型的木材缺陷图像分割算法[J].林业机械与木工设备,2018,46(4):44-47.CHENG Y Z,CAI Y F. Wood defect image segmentation algorithm based on fractional order CV model[J]. Forestry and Woodworking Equipment,2018,46(4):44-47.
[6]陈峥,石勇鹏,吉书鹏.一种改进的Otsu图像阈值分割算法[J].激光与红外,2012,42(5):584-588.Chen Z,Shi Y P,Ji S P. An improved threshold Segmentation for otsu images[J]. Laser and Infrared,2012,42(5):584-588.
[7]吴军,王龙龙.基于双鸟群混沌优化的otsu图像分割算法[J].微电子学与计算机,2018,35(12):119-124.WU J,WANG L L. Otsu image segmentation algorithm based on double Bird Group chaos optimization[J]. Microelectronics and Computer,2018,35(12):119-124.
[8]赵乃刚,邓景顺.粒子群优化算法综述[J].科技创新导报,2015,12(26):216-217.ZHAO N G,DENG J S. An overview of particle swarm optimization algorithm[J]. Scientific and Technological Innovation Guide,2015,12(26):216-217.
[9]李海楠,张学良.基于群体适应度方差的粒子群优化算法[J].计算机仿真,2007,24(5):158-161.LI H N,ZHANG X L. Particle swarm optimization algorithm based on variance of population fitness[J]. Computer Simulation,2007,24(5):158-161.
[10] MIRJALILI S. SCA:a sine cosine algorithm for solving optimization problems[J]. Knowledge-Based Systems,2016,96:120-133.
[11]刘勇,马良.转换参数非线性递减的正弦余弦算法[J].计算机工程与应用,2017,53(2):1-5.LIU Y,MA L. A sinusoidal cosine algorithm for nonlinear decrement of transfer parameters[J]. Computer Engineering and Application,2017,53(2):1-5.
[12]孙俊,宋佳,武小红.基于改进Otsu算法的生菜叶片图像分割方法[J].江苏大学学报,2018,39(2):179-184.SUN J,SONG J,WU X H. Segmentation of raw vegetable leaf image based on improved otsu algorithm[J]. Journal of Jiangsu University,2018,39(2):179-184.
[13]王裔,段会川. Otsu方法在多阈值图像分割中的应用[J].计算机工程与设计,2008,29(11):2844-2845.WANG Y,DUAN H C. Application of otsu method in multithreshold image segmentation[J]. Computer Engineering and Design,2008,29(11):2844-2845.
[14] OTSU N. A threshold selection method from gray-level histogram[J]. IEEE Transactions on System, Man, and Cybemetics,1979,9(1):62-66.
[15]何庆元,韩传久.基于粒子群算法的Otsu法图像阈值分割[J].桂林电子科技大学学报,2006,26(5):355-358.HE Q Y,HAN C J. Application of otsu method based on particle swarm optimization in image segmentation[J]. Journal of Guilin University of Electronic Science and Technology,2006,26(5):355-358.
[16]马文,周绍光,丁慧珍.基于遗传算法的Otsu法在图像分割中的应用[J].沿海企业与科技,2005,10(8):160-161.MA W,ZHOU S G,DING H Z. Application of otsu method based on genetic algorithm in image segmentation[J]. Coastal Enterprises and Science and Technology,2005,10(8):160-161.
[17] SHEIKH H R,SABIR M F,BOVIK A C. A statistical evaluation of recent full reference image quality assessment algorithms[J]. IEEE Transactions on Image Processing,2006,15(11):3441-3452.
[18] WANG Z,BOVIK A C,SHEIKH H R,et al. Image quality assessment:from error visibility to structural similarity[J]. IEEE transactions on image processing,2004,13(4):600-612.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |