点击次数：

论文类型：会议论文

发表时间：2016-01-01

收录刊物：CPCI-S

卷号：60

页面范围：1044-1049

关键字：Automated line-heating; image binarization; thresholding method

摘要：In this paper, a thresholding algorithm combining local and global region is proposed in order to perfect the vision of line-heating robot, segment foreground objects and background accurately, eliminate environment effects. The thresholding method combines the advantages of both local and global threshold algorithm. First, we detect the edge region of foreground objects with Canny operator and get the gray-level histogram of edge region. Next, according to the gray-level histogram, we calculate a segment threshold with OTSU method. Finally, the gray image is converted into binary image according to the segment threshold. The experimental results show that the thresholding method combining local and global region is effective on segmenting foreground objects and background. Our method could insure the shape integrity of foreground objects as well.