Factual Extraction with Choicest Retrieval in CBIR Using Angular Robust Techniques

Abstract

Content-Based Image Retrieval (CBIR) plays a vital role in various digital image processing fields such as image retrieval, classification, feature extraction, clustering, and indexing. In some cases, the high computational time as well the poor performance of similarity score makes the process of image retrieval an unsuitable one for CBIR having larger datasets. To overcome this trauma, the article presents a combined scheme to perform image retrieval and feature extraction using an eminent utilization of angular patterned wavelet Fourier descriptor with Randomized Robust Learning. With this prominent technique, both the local and the global feature descriptors are extracted to attain a maximum retrieval accuracy. Furthermore, the learning process is made prominent, owing to the unique extracted features. It makes the process of indexing similar images faster and the retrieval of relevant images with less computational time in an organized manner. Our proposed methodology is executed in MATLAB and is compared with the existing methodologies regarding retrieval accuracy and surfing time by fast indexed output.