边缘检测算法在不同分辨率图像中的性能研究  

Research on Performance of Edge Detection Operators in Image Processing under Different resolutions

作　　者：吴学明[1] 李灿平[2] 

机构地区：[1]成都理工大学信息管理学院,四川成都610059 [2]成都理工大学信息工程学院,四川成都610059

出　　处：《计算机测量与控制》2006年第2期166-169,共4页Computer Measurement &Control

摘　　要：边缘检测是图像处理、计算机视觉中的重要内容之一,是实现图像分割、特征提取和图像理解的基础,也是仍未得到圆满解决的研究课题;图像分辨率是衡量图像质量的重要参数,分辨率越高图像质量越好,存储空间也越大;重点讨论了数字图像边缘检测中常用的几种经典的边缘检测算法对图像在不同分辨率时的性能变化,以及图像在不同分辨率条件下加入噪声时的性能变化,分析了造成这些变化的主要原因;同时,通过实验得出了这些边缘检测算法的优缺点和一些经验性结论,以便在实际应用中更好地发挥它们的特长;实验证明,这些边缘检测算法对图像和加噪图像在不同分辨率下的性能及其变化存在明显的差异。Edge detection, which is still unresolved well till now, is a basic content of image processing and computer vision, and is also the base of image segmentation and image understanding, Image resolution is an important parameter which presents the quality of image. The higher is the resolution, the better the quality of the image, and the bigger the store space. Experiments prove that the performance of the edge-detection operators is different in image and noise contained image processing under different image resolutions. Therefore, the performance variations of several typical edge-detection operators in image and noise contained image processing under different resolution are discussed. Main reasons which cause these variations are also analyzed. Meanwhile, the advantages and disadvantages of these operators which result in the performance variations under different image resolutions and some experiential conclusions are drawn from the experiments.

关 键 词：边缘检测 算子 图像处理 图像分辨率 噪声 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]