查看文章

The study of Screen Content Image Quality Assessment (SCI-QA) is a new and interesting topic due to its excellent potential for instruction and optimization in various processing systems, it has been attractive recently. In this paper, we proposed a full reference quality assessment method for screen content images named Genetic Programming based Screen Content Image Quality (GP-SCIQ). The proposed method operates via a symbolic regression technique using Genetic Programming (GP). Hence, for predicting subject scores of images in datasets we combined the objective scores of a set of Image Quality Metrics (IQM). Two largest publicly available image databases (namely SICAD and SCID) are used for training and testing the predictive models, according the k-fold-cross-validation strategy. The performance of the proposed approach is evaluated, and several experiments are carried using four …