摘要
相对于1维Otsu算法,2维灰度级-梯度Otsu算法的抗噪声能力有所增强,但对受噪声影响较大的钢板表面缺陷图像的分割精度仍不高,且分割效率较低.为提高钢板表面缺陷图像分割的精度和效率,提出2维鲸鱼优化加权的灰度级-梯度Otsu算法.2维鲸鱼优化加权的灰度级-梯度Otsu算法能增大分布概率较低的阈值类间方差,减少背景与目标间差异对分割的影响.仿真实验结果表明,4种算法中,2维鲸鱼优化加权的灰度级-梯度Otsu算法的分割精度及效率均最高.
Compared with the one-dimensional Otsu algorithm,the two-dimensional grayscale-gradient Otsu algorithm had better anti-noise ability.However,the segmentation accuracy was still insufficient and the segmentation efficiency was low for the steel plate surface defect image which was greatly affected by noise.In order to improve the accuracy and efficiency of plate surface defect image segmentation,the two-dimensional weighted grayscale-gradient Otsu algorithm with whale optimization was proposed.The two-dimensional weighted grayscale-gradient Otsu algorithm with whale optimization increased the inter-class variance of threshold with lower distribution probability,and reduced the influence of the large difference between the background and the target on segmentation.The simulation results showed that the segmentation accuracy and efficiency of the two-dimensional weighted grayscale-gradient Otsu algorithm with whale optimization were the highest among the four algorithms.
作者
张东洋
杨永辉
储茂祥
邓鑫
ZHANG Dongyang;YANG Yonghui;CHU Maoxiang;DENG Xin(Institute of Electronics and Information Engineering, University of Science and Technology Liaoning, Anshan 114051, China)
出处
《安徽大学学报(自然科学版)》
CAS
北大核心
2020年第3期72-77,共6页
Journal of Anhui University(Natural Science Edition)
基金
国家自然科学基金资助项目(51674140,71771112)
辽宁省科学技术基金资助项目(20180550067)
辽宁省自然科学基金资助项目(20170540472)
辽宁省高等学校基本科研项目(2017LNQN11)。
关键词
钢板表面缺陷
2维加权Otsu算法
鲸鱼优化算法
精度
效率
steel plate surface defects
weighted two-dimensional Otsu algorithm
whale optimization algorithm
accuracy
efficiency