In this paper, an efficient method for Content Based Image Retrieval (CBIR) to retrieve the images of diabetic retinopathy (DR) is proposed. The methodology involves inter-plane relationship between pixel intensities and feature reduction. Key pixels are selected from an edgy image and are used for computing the intensities in inter-plane relationship. By using the selected point as a center pixel the Local Binary Patterns (LBPs) are computed. Our approach enhanced the results by compressing the size of the resultant metrics. Feature reduction is performed by using the Random Bin Selection method. The experiments conducted on the STARE dataset indicates an increment of 42.04% in the average precision rate compared to the existing method.
L. Suresh, Chandran, S., Vijayan, D., and E. R. Vimina, “An Efficient method to Retrieve Diabetic Retinopathy Images using CBIR Technique”, in 2020 Fourth International Conference on Computing Methodologies and Communication (ICCMC), Erode, India, 2020.