基于掩膜与仿中值滤波的边缘检测处理

Partitioning Algorithm Based on Mask and Imitated Median Filtering and Its Application

针对医学图像特殊而又复杂的模糊边缘以及难以区分的背景噪声,提出一种基于掩膜理论区分图像边缘信息的区分算法.通过区域对比度、图像像素特征和有限区域中像素均值等相关信息,利用先验知识和仿中值滤波的方法得到较为清晰的边缘信息.用Matlab软件对提出的方法进行模拟,详细分析了掩膜对比的理论模型以及仿中值滤波对于区分效果的影响,验证了该方法的可行性和正确性.实验证明,基于掩膜与仿中值滤波的区分算法和传统的边缘检测方法相比,在辨别相关边缘信息方面具有较好的效果,并在医学图像处理中具有一定的实用价值.

Considering the special and complicated fuzzy edges and the difficulty to distinguish the background noise in medical images,apartitioning algorithm based on the mask theory was proposed to distinguish the marginal information.The edge information was given by the area contrast,the characteristic of image pixel and the mean value of pixel on the enclosed area.In addition,more clear edge information was further gained by using the prior knowledge and imitation median filtering.The proposed method was implemented by meas of the software MATLAB.The influences of the theoretical model of the mask contrast and the imitated median filtering on the distinguishing effect were minutely analyzed and the method was verified to be feasible and accurate.The experiments show that the partitioning algorithm based on the masking theory and imitated median filtering can achieve better results on distinguishing marginal information compared with the traditional edge detection method,and has certain practical value in medical image processings.