亚像素渲染
像素
计算机视觉
人工智能
边缘检测
分割
GSM演进的增强数据速率
计算机科学
图像分割
图像处理
数学
图像(数学)
摘要
Aiming at the poor positioning accuracy and the weak ability to resist noise in traditional edge detection methods,a new approach was proposed using the one dimension gray moment to determine subpixel edge location. Firstly,the position range of image edge would be obtained by using threshold segmentation. Secondly,coarse positioning should be done by expanding pixels along the threshold segmentation boundary points. At last,fine positioning should be done by expanding pixels along the coarse localization boundary points until the difference of the two fine positioning results are not greater than 0. 01 pixel. Compared with other subpixel edge location methods,edge positioning accuracy obtained by the proposed method is better. The experiment results showed that the accuracycan reach 0. 05 pixel( standard image) and 0. 14 pixel( actual image) respectively.
科研通智能强力驱动
Strongly Powered by AbleSci AI