基于边缘检测器组合的密集边缘检测

Composite Edge Detectors for Closely Spaced Edge Detection

本文提出了基于多个边缘检测器组合的密集边缘检测方法 ,并从理论上阐述了此方法的性质 .在实验中 ,采取了组合中值混合滤波器、Teboul变分法和最优的 Gabor滤波器的方式 ,随后用组合的边缘检测器对人脑横切片图和取自 Vis Tex数据库的图像进行分割 .实验结果表明此方法是有效的 .

In this paper, a detecting method for closely spaced edge detection based on multiple edge-detector combination is proposed, and the properties of composite edge detectors are also elaborated in theory. In the experiment, a linear-median hybrid filter, Teboulian calculus of variations and optimized Gabor filters are combined to segment and detect transverse slice image of human head and images taken from Vis Tex-database. The experimental results show that this method is effective.