A Content Based Image Retrieval Approach Based On Color and Shape
Khalid El Asnaoui, Youness Chawki, Brahim Aksasse, Mohammed Ouanan · International journal of tomography and simulation · 2016
The problem that we consider in this paper is about the indexing images in a large database. In this context, we propose a new technique for indexing based on the combination of color and shape. This is an improvement of the 2-D histogram method with statistical moments by integrating the shape using Angular Radial Transform (ART) and by applying distributed computation to research image. This model reduces the searching time and provides a rapid response. Several experiments were performed using two different databases: Wang and Coil-100. In addition, the results have demonstrated the effectiveness and the robustness of the new method to the two databases with the best rate of average precision.