Abstract
In medical imaging lossless coding is commonly used to avoid the risk of misinterpretation. However, some visual distortions are invisible by Human Vision System (HVS) if errors are below Just Noticeable Distortions (JND). Actually JND refers to the maximum distortion that the human visual system (HVS) cannot perceive. JND depends on many factors like average brightness, contrast,
colorfulness, temporal activity etc.
In this paper we use the well-known block-based model for JND estimation, the method is dependent on average brightness only (other factors are ignored). Upon getting the image JND, we apply the near-lossless JPEG-LS coding with the controlled distortion below or equal to the JND.
As expected, JND-based near-lossless coding provides higher compression ratios without compromising visual quality.
Our method targets mHealth applications as Remote Diagnosis and Care (incl. using tele-operated robots), Second Opinion Provision, Medical Education and Home Monitoring.
JND is higher at regions of low and high brightness
The full paper is here
23+ years’ programming and theoretical experience in the computer science fields such as video compression, media streaming and artificial intelligence (co-author of several papers and patents).
the author is looking for new job, my resume
