面向钢丝绳缺陷检测的图像预处理方法

Image Preprocessing Method for Detection of the Wire Rope Defects

钢丝绳在使用过程中可能会出现断丝、蜷曲变形,甚至瞬间断裂造成严重的安全事故。传统依靠人工目检的方式正在被新的检测技术替换。图像检测方式的准确性和便利性普遍得到行业认可,但成像设备受到光线、成本、安装距离等因素限制,无法提供到后级处理单元足够的图像信噪比,带来误检、漏检的可能。因此,本文将C-BM3D去噪算法引入到钢丝绳图像检测设备前端,并针对其他几种去噪算法在钢丝绳检测领域的实际效果进行了对比分析。依据PSNR进行性能评定,实验结果表明,C-BM3D算法的引入对实际钢丝绳图像检测的准确性带来进一步的提高。

In recent years,the accuracy and convenience of image detection methods are generally recognized by the industry.However,due to the limitation of light,cost,installation distance and other factors,imaging equipment cannot provide enough image signal-to-noise ratio to the post-processing unit,which brings the possibility of false detection and missing detection.In this paper,c-bm3d denoising algorithm is introduced to the front-end of wire rope image detection equipment,and the actual effect of several other denoising algorithms in the field of wire rope detection is compared and analyzed.The experimental results show that the introduction of the algorithm can further improve the accuracy of the actual wire rope image detection.