三阶差分运算与传统图像边缘检测算法的对比分析  

Comparative analysis for third order difference arithmetic with typical image edge detection algorithms

作　　者：向应霞 丁岳伟[1] Xiang Yingxia;Ding Yuewei(School of Optical-Electrical and Computer Engineering,University o[Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学光电信息与计算机科学学院

出　　处：《电子测量技术》2018年第23期49-54,共6页Electronic Measurement Technology

摘　　要：图像边缘检测技术是计算机视觉领域的重要内容,且一直是图像处理与分析领域的研究热点,研究者们提出了多种经典的边缘检测算法。为了在以后的不同应用中选取最合适的边缘检测算法类型,本文采用对比分析法,对近几年提出来的三阶差分边缘检测算法进行了研究,在基于不同的图片格式基础上,与其他传统边缘检测算法进行对比,比较了图像熵值,得出算法与图片格式具有相关性、各算法对噪声不同程度的敏感性、三阶差分运算对图片格式具有选择性、不能用图像熵作为评价边缘检测算法性能的结论,便于对不同场景选择何种图片存储格式及何种边缘检测算法作出选择。Image edge detection technology has become an important part in the field of computer vision and it has always been a research hotspot in the field of image processing and analysis,many researchers also proposed a variety of edge detection algorithms.In order to select the most appropriate type of edge detection algorithm in different situations, according to the newly proposed third order difference edge detection algorithm and compared with other traditional edge detection algorithms and image entropy, the paper drawing the conclusions that edge detection algorithms are associated with image formats, different edge detection algorithms are sensitive to noise to some extent and image entropy can’t be an evaluation as the edge detection algorithm performance. These conclusions can be contribute to choose an image storage format in different scenes and an optimal edge detection algorithm.

关 键 词：图像处理 边缘检测 三阶差分 图像熵 图片格式 

分 类 号：TN911.73[电子电信—通信与信息系统]