Accelerating of color moments and texture features extraction using GPU based parallel computing

Image retrieval tools can assist people in making efficient use of digital image collections; also it has become imperative to find efficient methods for the retrieval of these images. Most image processing algorithms are inherently parallel, so multithreading processors are suitable in such applications. In very big image databases, image processing takes very long time for run on a single core processor because of single thread execution of algorithms. GPU is more common in most image processing applications due to multithread execution of algorithms, programmability and low cost. In this paper we implement color moments and texture based image retrieval (entropy, standard deviation and local range) in parallel using CUDA programming model to run on GPUs. These features are applied to search images from a database which are similar to a query image. We evaluated our retrieval system using recall, precision, and average precision measures. Experimental results showed that parallel implementation led to an average speed up of 144.67×over the serial implementation when running on a NVIDIA GPU GeForce GT610M. Also the average precision and the average recall of proposed method are 61.968% and 55% respectively.