摘 要：数字图像处理技术是信息科学中近几十年来发展最为迅速的学术之一。图像边缘是图像最基本的一种特征，扮演着一个重要的角色，在图像识别，图像分割，图像增强以及图像压缩等领域中有较为广泛的应用，其目的就是精确定位缺陷，同时更好地抑制噪声。图像边缘提取的手段多种多样，本文主要通过在 MATLAB平台上做图像的缺陷检测 ，把彩色图像从RGB空间转换到XYZ空间来分析。通过中值滤波，Retinex理论的图像增强，方差算法来实现静态图像的缺陷检测。

关键词：数字图像；图像处理；缺陷检测

Abstract：The digital image processing technology in information science is one of the fastest growing disciplines in recent decades. Edge is a basic feature in image and the edge of the image plays an important role.There is more widely used in the image recognition, image segmentation, image enhancement and image compression and other areas. The purpose is to accurately locate the defect, and remove noise . There are many means of image defect extraction. In this paper different operators are used to achieve a static image defect detection in MATLAB programming language . we can transform RGB image to XYZ image ，to process image by using Median filter and image enhancement of the Retinex theory .And Static image can be detected through the Variance algorithm .  

Keywords : Digital images；Image edge；Defect detection