摘要
针对现有的边缘检测算法在应对边缘灰度变化非常低且有划痕情况下的工件图像噪声消除时,所产生的图片边缘信息丢失,使得边缘提取时因丢失大量边缘信息且难以消除工件表面划痕干扰的问题,提出一种将改进的高斯滤波和中值滤波相结合的滤波方法,在有效消除大量高斯噪声和椒盐噪声的同时还有效地保留了大量的边缘信息;此外,为了实现梯度幅值和方向的增强计算,使用Sobel算法的卷积核对图像进行卷积;再用非极大值抑制法对边缘进行细化;最后采用Otsu算法进行自适应高低阈值的选取,完成边缘检测。实验结果表明,改进的边缘检测方法与传统方法相比,消除噪声更好,边缘更连续,图像质量更好,算法自适应性更高,有良好的改进效果。
Aiming at image edge information loss in response to artifact image noise elimination with very low edge grayscale changes and scratches,it is difficult to eliminate the problem of the loss of edge information,A filtering method combining improved Gaussian and median filtering is proposed.While effectively eliminating a large amount of Gaussian noise and pretzel noise,it also effectively retains a large amount of edge information.Besides,to achieve the enhanced computation of the gradient amplitude and direction,images were convoluted using the convolution of the Sobel operator.The edges were refined again by non-maximum inhibition.Finally,the Otsu algorithm was used to select the adaptive high and low threshold values for completing the edge detection.The experimental results show that the improved edge detection method has better noise elimination,more continuous edges,better image quality,higher algorithm adaptability,and improved effect compared with the traditional methods.
作者
吴云飞
柏受军
江明
徐印赟
WU Yunfei;BAI Shoujun;JIANG Ming;XU Yinyun(Key Laboratory of Advanced Perception and Intelligent Control of High-end Equipment of Ministry of Education,Anhui Polytechnic University,Wuhu 241000,China;School of Electrical Engineering,Anhui Polytechnic University,Wuhu 241000,China)
出处
《安徽工程大学学报》
CAS
2022年第4期32-40,45,共10页
Journal of Anhui Polytechnic University
