Image distortion analysis and classification scheme using a neural approach
Aladine Chetouani, Azeddine Beghdadi, Mohamed A. Deriche · 2010
Generally, IQMs are not able to well predict the image quality for all degradations. Indeed, well performance could be obtained for a given degradation and poor results for others. This is essentially due to the fact that the efficiency of IQMs depends highly on the degradation specificity. To overcome this limitation, we propose to first identify the type of degradation before measuring the quality of the image. The degradation classification is performed using an Artificial Neural Networks (ANN). Then, the most appropriate IQM could be used to estimate the image quality. The obtained results in terms of classification prove the efficiency of the proposed method. This scheme could provide a powerful image distortions recognition and classification that could be used in a image quality assessment system.