基于相位一致模型的边缘检测

Image feature detection based on phase congruency model

边缘检测在图像处理和理解中占有重要地位,传统的边缘检测是基于梯度算子,而相位一致方法是基于傅里叶变换理论的一种图像特征检测方法。其原理是指图像特征总发生在相位的最大叠合处。通过计算局部能量,来度量其图像各个位置的相位一致值。利用log Gabor滤波器组实现相位一致模型。为了检验相位一致模型在边缘检测中的有效性,利用该方法对工业图像和遥感图像进行边缘检测实验,并将该方法与canny算子和sobel算子进行对比分析。结果表明相位一致方法可以有效地提取边缘特征。

The algorithm of phase congruency developed from Fourier transform is applied to the image edge detection. The theory on phase congruency is that image edges always occur at points where the phases of harmonic components come to the maximum congruency. The edge detection is completed by constructing the local energy model. The log Gabor function for detecting the image edge features is introduced and compared with the canny and sobel operators. The proposed algorithm of Phase Congruency is implemented into the industry image and remotely-sensed image. The results show that the algorithm is effective to detect image edges.