Abstract

The restoration of an image with blocking artifacts due to compression at low bit rates is a challenging task and blocking artifact measurement algorithms have an important role to play in the computer vision field. An artifacts removal technique is an important step towards the reliability and security of image processing area that delivers a better understanding in many applications like pattern recognition, object classification, surveillance system and many more. We know that the removal of art objects is a scientific method used to provide better image analysis and for this purpose many methods of removal of art objects were already made by researchers during the processing of images such as line, motion, pattern, and hair. But in availability of group of artifacts in an image, they do not achieve an acceptable result. In this research, we proposed a comparative framework for blocking artifacts removal of digital images using hybrid mechanism. The main contribution of this research is developing a new neuro-fuzzy system-based hybrid artifacts removal mechanism to achieve better blocking artifacts efficiency. To remove artifact from an image the proposed framework has its own impact in quality parameters such as Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), and Structural Similarity (SSIM) with the execution time. At last, the performance parameters of proposed framework is compare for all five techniques such as line, motion, pattern, hair and combination of all with each other and we observed that the achieved results justify the proposed hybrid artifact removal method in the field of image processing.