Image Deblurring Via Total Variation Based Structured Sparse Model Selection

Retina imaging technology is an effective control method for early diagnosis and early treatment of the diabetic retinopathy. In this paper, a fast robust inverse diffusion equation combining a blockwise filtering is presented to detect and evaluate diabetic retinopathy using retinal image enhancement. A flux corrected transport technique is used to control diffusion flux adaptively, which eliminates overshoots inherent in the Laplacian operation. Feature preserving denoising by the blockwise filtering ensures a robust enhancement of noisy and blurry retinal images. Experimental results demonstrate that this algorithm can enhance important details of retinal image data effectively, affording an opportunity for better medical interpretation and subsequent processing. Main contributions: 1. Fractional inverse diffusion. In order to make full use of adjacent information of image pixels, we propose an inverse diffusion equation with the fractional differential orders controlled by local image features. The enhanced adaptivity in image sharpening affords a remarkable enhancement of main image features, while reducing the background and noise interferences. 2. Fast blockwise filtering with noise analysis. With the help of the principal component analysis of image noise, the parameters of image noise are estimated. Then, a fast blockwise filtering is presented based on the collaborative filtering of similar image blocks. A new similarity between image blocks is defined to guarantee effective sparse processing. 11:15 C10-3(invited) A Continuous Random Walker Model with Explicit Coherence Regularization for Image Segmentation