An efficient content based image retrieval based on speeded up robust features (SURF) with optimization technique

Abstract

The content-based image retrieval (CBIR) plays an essential role in the development of social networking medium and image retrieval process. To improve the retrieval performance and decrease the margin between the visual features the CBIR process is used. In the content-based image retrieval, the feature of the query image and the database image are automatically matched. The features like color, texture, shape and histogram etc. are used to retrieve the similar in the large database. In the proposed work, an efficient content-based image retrieval system based on speeded up robust features (SURF) with optimization technique is presented. The SURF feature descriptor technique is used to extract the feature from the query image and basically the SURF technique appropriate feature extraction technique for the color image. In existing feature extraction technique, a limited number of pixels partitions are available so they cannot extract the whole color feature from the color image. So in proposed work, we use the combination of SURF technique with the genetic algorithm to improve the performance parameters like precision, recall, f-measure, and accuracy. After the simulation of proposed CBIR system, we concluded the accuracy more than 98% with a better precision and recall value. The content-based image retrieval system based on SURF with optimization technique is implemented using Image Processing Toolbox within the MATLAB Software.