Face Image Retrieval Using Genetic Algorithm and Bags of Pixels

摘要

This paper presents a method for image retrieval using genetic algorithm and bags of pixel. This technique makes it possible to select important parts of image for retrieval. Closed space of pixels takes advantage of automatic conformity approximation and indicates the increase of meaningful linear variations such as morphing and transitions. Selecting bags of pixels and PCA, we can retrieve images from the database. For image retrieval, we can use the best color features of the retrieved image which are selected based on genetic algorithm. In this method, the massive of saved information becomes less and the speed of retrieval operation increases. The accuracy of this method reaches to 89.8% and linear variations and transitions have no effects on retrieval operation.