计量学报2024，Vol.45Issue(8)：1200-1208,9.DOI:10.3969/j.issn.1000-1158.2024.08.16
基于自适应BM3D的绝缘子缺陷检测图像降噪方法
Image Denoising Method for Insulator Defect Detection Based on Adaptive BM3D
摘要
Abstract
Due to the influence of the shooting environment,the acquired image is often mixed with noise,which easily affects the accuracy of insulator defect detection.To solve this problem,an adaptive BM3 D noise reduction method is proposed.First,a noise estimation algorithm based on the statistical relationship between the noise level and the eigenvalue of the image block covariance matrix is introduced to solve the problem that the original BM3D algorithm needs prior knowledge.Second,taking the peak signal to noise ratio(PSNR)as the objective function,the optimal parameters of the insulator image under each noise intensity are obtained by quantum genetic algorithm,including the hard threshold parameter and the distance threshold in the basic estimation,and the distance threshold in the final estimation.Finally,taking the noise intensity as the independent variable,the fitting curve of the above three parameters are obtained by polynomial fitting,so as to obtain the optimal parameters combination of the algorithm under each noise intensity,and realize the rapid parameter adaptation of BM3D algorithm under different noise levels.The results of comparative experiments indicate that the proposed method is superior to other methods in terms of visual and objective evaluation indexes.When the noise standard deviation is 25,the proposed method improves PSNR,structural similarity(SSIM)and edge preserve index(EPI)compared with the original BM3D algorithm,especially the EPI value increases by nearly 20％.While improving the noise reduction effect,it can retain more edge details,which is helpful to improve the effect of subsequent insulator identification and defect detection.关键词
电学计量/绝缘子缺陷检测/图像识别/图像降噪/BM3D算法/结构相似性Key words
electrical measurement/insulator defect detection/image recognition/image denoising/BM3D/SSIM分类
通用工业技术引用本文复制引用
时培明,袁群贸,许学方,阚俊明..基于自适应BM3D的绝缘子缺陷检测图像降噪方法[J].计量学报,2024,45(8):1200-1208,9.基金项目
河北省自然科学基金(E2020203147 （）
E2022203093) （）
中央引导地方科技发展资金项目(216Z4301G) （216Z4301G）
