New Color and Texture Features Coding Method Combined to the Simulated Annealing Algorithm for Content Based Image Retrieval

Content based image retrieval is a method that can find similar images to a query image in a large database. This technique is based on the perceptible contents and visual features of the image. In this work, we implemented a new method of coding the visual features of the image based on strings. The aim of this work is to improve our previous method by reducing the retrieval time and increasing the number of results by improving the two measures: precision and recall. Therefore, from a query image we extract the color signature. Then, we create the gray level co-occurrence matrix to extract four texture features. Meanwhile, we develop our previous method which is color string coding while combining color and texture features in the purpose of creating an efficient codification technique which improve enormously the size of coding the image descriptors. Finally, the simulated annealing is implemented as a low complexity meta-heuristic algorithm for image retrieval.