An Approach to JPEG Image Artifact Reduction based on Type-2 Fuzzy Hybrid System

  • Nadeem Ahmad, Zainul Abdin Jaffery, Shaheen Khan, Irshad

Abstract

The rapid growth in multimedia data has put considerable pressure on computer hardware and memory management system to develop an energy-efficient, low bit-rate image codec for information transmission and retrieval for the band-limited channel. It has noted, at low bit-rate, JPEG suffers from visual distortion in the image. To reduced the artifact problems from JPEG  codec, this paper proposed a Type-2 fuzzy-based image filtering method, which is suitable to suppress the visual distortion issue raised in JPEG image coding. This proposed method deals in reducing the three types of JPEG image artifacts: Blocking, blurring of edges, and aliasing. The whole process has divided into two steps to reduce the JPEG image visual distortion effects. In the first step, boundary smoothing and filtering techniques have used to remove the blocking artifacts. In contrast, in a second step, blurring and aliasing effects around the edges have removed through a Type-2 fuzzy based algorithm. The artifacts reduction efficiency of the proposed method has been compared with other existing JPEG images, artifact removal methods. The visual quality assessment of JPEG images has decided by computing the Peak Signal to Noise Ratio (PSNR) of each technique. The proposed method outperforms the other existing method in terms of PSNR improvement in the range of 0.22 to 1.10 dB

Published
2020-06-01
How to Cite
Nadeem Ahmad, Zainul Abdin Jaffery, Shaheen Khan, Irshad. (2020). An Approach to JPEG Image Artifact Reduction based on Type-2 Fuzzy Hybrid System. International Journal of Advanced Science and Technology, 29(7s), 5936-5949. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/26518