【摘要】 本文提出一种新的利用5×5窗口的边缘检测方法。它可以明显地抑制噪声,并且克服了5×5窗口使被检边缘宽度增加的副作用。另外,针对图象成象时曝光不足的区域,对比度下降的缺陷,对不同灰度背景的区域,采用不同的判别门槛,因此能够正确识别不同灰度背景的物体边缘。更多还原

【Abstract】 This paper presents a new edge detector using 5x5 window, which can reduce the noise effect efficiently but not increase the width of the detected edge which always happened,, in case of using 5x5 window edge detector. Besides that, in order to handle the problem that the contrast reduced in the dark region is caused by underexposure, this detector uses a auto-changeable threshold, so that it can detect the edge in regions of different grey background correctly.更多还原