This is a brief introduction of our proposed JND profile with totally new design philosophy based on KLT. You can change our program as you like and use it for academic, but please refer to its original source and cite our paper.
Update: we provide python extension of this JND profile KLT_JND.py for research usage!
Just noticeable difference (JND) of natural images refers to the maximum pixel intensity change magnitude that typical human visual system (HVS) cannot perceive. Existing efforts on JND estimation mainly dedicate to modeling the diverse masking effects in either/both spatial or/and frequency domains, and then fusing them into an overall JND estimate. In this work, we turn to a dramatically different way to address this problem with a top-down design philosophy. Instead of explicitly formulating and fusing different masking effects in a bottom-up way, the proposed JND estimation model dedicates to first predicting a critical perceptual lossless (CPL) counterpart of the original image and then calculating the difference map between the original image and the predicted CPL image as the JND map. We conduct subjective experiments to determine the critical points of 500 images and find that the distribution of cumulative normalized KLT coefficient energy values over all 500 images at these critical points can be well characterized by a Weibull distribution. Given a testing image, its corresponding critical point is determined by a simple weighted average scheme where the weights are determined by a fitted Weibull distribution function. The performance of the proposed JND model is evaluated explicitly with direct JND prediction and implicitly with three applications including JND-guided noise injection, JND-guided image compression, and distortion detection and discrimination. Experimental results have demonstrated that promising performance of the proposed JND model.
You can download our proposed software in this project page directly.
Matlab R2019a
Important Note: other versions may lead to some errors.
If you have any questions of this repo or our paper, please feel free to contact with the authors: jiangqiuping@nbu.edu.cn, zhentaoliu0319@163.com.
If you find this work is useful for you, please cite our paper.
@ARTICLE{KLT-JND,
author={Jiang, Qiuping and Liu, Zhentao and Wang, Shiqi and Shao, Feng and Lin, Weisi},
journal={IEEE Transactions on Image Processing},
title={Toward Top-Down Just Noticeable Difference Estimation of Natural Images},
year={2022},
volume={31},
number={},
pages={3697-3712},
doi={10.1109/TIP.2022.3174398}}