Visual repercussions of the noise induced errors in the communication channel on compressed images

Abstract

This research analyzes the susceptibility to errors in the communication channel of the two standards in image compression - JPEG and JPEG 2000. Errors are introduced to the bit stream via channel noise, i.e. Gauss, Poisson, uniform noise and their influence to image quality is measured by objective and subjective image quality assessment methods. A novel compression method based on partial differential equations is introduced as a direct contender to two before mentioned standards.