摘要
边缘是图像最基本的特征,它包含对人类视觉和机器视觉有价值的物体边缘信息,边缘是图像中特性(如像素灰度、纹理等)分布不连续处,图像周围特性有阶跃变化或屋脊变幻的那些像素的集合。由于采集、光照等原因,尿液红细胞所在的区域包含非常丰富的细节(边缘特征很明显),本文利用Sobel算法,对比Roberts算法,然后对图像进行细化,以达到对红细胞进行边缘检测的目的。
Edge is the most basic feature of image, It contains the object edge information of value to the human vision and machine vision, Edge is the image characteristics （ such as pixel gray level, texture, etc.） distribution is not continuous, Those pixels set step change or roof changing around the image characteristics. As a result of the acquisition, light and other reasons, Urine red blood cell area contains very rich details （ edge features clear ）, In this paper, by using the Sobel algorithm, Comparison of Roberts algorithm, Then the refinement of the image,The goal of edge detection of erythrocyte to achieve.
出处
《软件》
2013年第1期158-159,共2页
Software
