摘要： 传统SUSAN算法在提取图像边缘时，会出现漏检现象，且所提取的边缘较粗。为此，运用计算最大类间方差的方法自适应地选取双阈值，取代传统算法中人工设定的单阈值，采用多方向局部非极大值抑制方法进行改进，提出一种新的SUSAN边缘检测算法，并将其应用于遥感图像的边缘提取。实验结果表明，该算法能够有效提高边缘定位精度，降低漏检率，使边缘更细致光滑。
关键词:
图像处理,
边缘检测,
SUSAN算法,
遥感图像,
自适应双阈值,
最大类间方差法,
局部非极大值抑制
Abstract: The traditional SUSAN edge detection algorithm is lack of detecting fine edge. Adaptively dual-threshold algorithm using Otsu’s method is selected to replace the traditional manual single threshold. And multi-direction local non-maxima suppression method is proposed to improve the SUSAN edge detection algorithm. The new algorithm is applied to remote sensing image. Experimental results show the new algorithm effectively improves the accuracy and reduce the missing rate, and the edge is more detailed and continuous.
Key words:
image processing,
edge detection,
SUSAN algorithm,
remote sensing image,
adaptive dual-threshold,
Otsu’s method,
local non- maximum suppression
中图分类号:
钟顺虹, 何建农. 基于自适应双阈值的SUSAN算法?[J]. 计算机工程, 2012, 38(3): 206-208,211.
ZHONG Shun-Gong, HE Jian-Nong. SUSAN Algorithm Based on Adaptive Dual-threshold[J]. Computer Engineering, 2012, 38(3): 206-208,211.