Image Filtering Method Based On Image Prior Knowledge

Digital images collected by image sensors are generally contaminated in the process of data acquisition, digitization, compression and transmission. Many types of noises or degradations may be included in the image. Thus, it is necessary to take image filtering before image data is analyzed. Image filtering methods are generally categorized into two basic approaches: one used in the cases when we lack prior knowledge of the specific input image, and another used in the cases when we know some factors causing image noises or degradations, or have some prior knowledge about the contents of the images, thus the prior knowledge is utilized for improving image filtering performance. Many image filtering methods have been proposed using the histogram or estimated histogram of original image as the priori knowledge. As an important image feature, edge always contains significant information in the scene, it can be used as a good expression of image prior knowledge.In this paper we present a novel multi-mask filtering method based on image edge knowledge. Considering the space-time summation and fire-capture characteristics of pulse-coupled neural network (PCNN), this method can detect the edge of noisy image effectively. The detected edge can be treated as the prior knowledge of original image to improve the subsequent image filtering performance. We also develop a median-based Gaussian weighted filter (MGWF) that is effective for image edge filtering. Through experiments on benchmark test images from Granada University, we demonstrate that compared with some previously published methods, our method has better performance in terms of visual effect and mean squared error.