摘要
在采集电路板刀具磨损图像过程中,会因光照角度的变化导致图像细节不足,而常规图像细节增强方法在处理过程中存在噪声混叠的情况,实时性不强。针对这一问题提出基于单尺度Retinex的电路板刀具磨损图像细节增强方法,根据目标图像的灰度值变化来定义灰度色调函数,在构建图像处理模型的基础上,采用引导滤波函数处理目标图像模型,去除图像噪声干扰后,使用高斯函数获得图像内的照射分量和反射分量,再利用单尺度Retinex方法剔除照射分量,增强电路板刀具磨损图像的细节。实验结果表明:设计的基于单尺度Retinex的图像细节增强方法运行时间短,增强效果好,处理后的图像质量高。
In the process of tool wear image acquisition of circuit board, the image detail is insufficient due to the change of illumination angle, and there is noise overlap by conventional image detail enhancement method during processing, which leads to the lack of real-time enhancement method.A single-scale Retinex-based detail enhancement method for circuit board tool wear images is proposed.The gray tone function is defined by the change of the gray value of the target image, on the basis of the construction of the image processing model, the target image model is processed by using a guided filter function, and after removing the image noise interference, the irradiation component and reflection component within the image are obtained by Gaussian function, and then the irradiation component is removed by the single-scale Retinex method to enhance the details of the circuit board tool wear image.The experimental results show that the designed single-scale Retinex-based image detail enhancement method has short running time, good enhancement effect, and high quality of the processed images.
作者
杨晶晶
欧冰
李海
Yang Jingjing;Ou Bing;Li Hai(Zhongshan Polytechnic,Zhongshan,Guangdong 528400,China)
出处
《工具技术》
北大核心
2022年第6期117-120,共4页
Tool Engineering
基金
广东省教育厅科研项目(2020KTSCX328)。