Ranking images based on shape features extracted using CBIR Techniques

Abstract

Content-Based Image Retrieval (CBIR) has emerged as a powerful technique for efficiently searching and retrieving images from large datasets based on visual features rather than textual metadata. This paper focuses on ranking images based on shape features extracted using CBIR techniques, employing methods such as boundary moments, complex coordinates, curvature scale space, intersection point map and merging methods. Each of these methods is utilized to analyze and extract distinctive shape features from images, which are then used for similarity matching and ranking. Ranking images based on shape features plays a vital role in Content-Based Image Retrieval (CBIR), where the goal is to efficiently retrieve and rank images based on their visual similarity to a query image. This study explores the extraction and utilization of shape-based visual features such as contours, geometric properties, and region-based descriptors for ranking images. A similarity measure is used to compare the feature vectors of the query image against those in a database. The retrieved images are ranked in descending order of similarity, ensuring relevance and precision in retrieval.