Image Quality Assessment In Big Data Environment

The visual information is the most important way for people to get information. Along with the rapid development of information technology, electronic technology and computer science, so the number of video applications and services is also increasing. But in the processing of digital images, distortions are inevitable. To maintain and improve the quality of the image, it is important to identify and quantify image quality. Big data bring both challenges and opportunities to image quality assessment. The challenges are the speed and real-time computing is difficult to ensure. The opportunities are vast amounts of data bring a wealth of information that can be used.This thesis reviews the development of image quality assessment algorithms. Image quality assessment has been described from the aspects of signal to noise, structural information and statistical characteristics of images.From characteristics of human visual system this thesis proposed these assumptions:(1) Texture affected people assess the quality of images.(2) When the person carrying out quality assessment used in the past experience. From the characteristics of big data this thesis propose these ideas:(1)parallel computing environment requires images to be divided into blocks(2) big data environments can find similar images to simulate the human experience. An image quality assessment framework is built based on the idea presents above.Experiments have been conduct to verify the ideas proposed. The first experiment verified Gabor texture and block size’s affect on image quality assessment. The second experiment verified similar image quality assessment use LIVE mobile video database as similar images to train the algorithm.Finally, a summary of the full text and prospects for future work has been made.