基于Sobel算子含噪低对比度图像的边缘检测方法  

Approach for Noise and Low-contrast Image Edge Detection Based on Sobel Operator

作　　者：郑敏[1] 王有熙[1] 税冬东[1] 

机构地区：[1]武警工程学院,陕西西安710086

出　　处：《石河子大学学报（自然科学版）》2008年第1期117-119,共3页Journal of Shihezi University(Natural Science)

摘　　要：在分析噪声图像、低对比度图像的各种边缘检测方法及Sobel算子检测原理的基础上,提出了一种以Sobel算子为基础,应用直方图均衡化和二值化的改进算法。经实验证明,改进过的算法对含噪低对比度图像具有较好的检测效果,与传统的Sobel、Canny算法相比,具有边缘连续、清晰的优越性。Based on the analysis of approaches for noise, low-contrast image edge detection and Sobel operator detection principle, this paper presents an improved algorithm based on Sobel operator. This algorithm employs the theory of linear equal histogram and binary processing in edge detection. Experimentresults are obtained by applying the improved edge detector on the noisy and low-contrast images, which prove to be more efficient than the Sobel and the Canny edge detectors.

关 键 词：边缘检测 SOBEL算子 直方图均衡化 含噪低对比度图像 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]