A Detail-maintained Tone Mapping Algorithm Based on Weighted Least Squares Filter
Hong Cheng, Yuxin Wei, Jianlin Zhang, Zhiyong Xu · 2019 IEEE 2nd International Conference on Automation, Electronics and Electrical Engineering (AUTEEE) · 2019
High dynamic range (HDR) images require tone mapping to be displayed properly on lower dynamic range devices. This paper presents a detail-maintained tone mapping algorithm based on weighted least squares (WLS) filter, which is an improvement of the ATT algorithm. Our new algorithm can overcome the shortcomings of the ATT algorithm, which may lose the details of the HDR picture and cause color distortion for some colorful pictures simultaneously. We firstly decompose the input HDR image into a base layer and a detail layer by the WLS filter. Then for the base layer, its dynamic range is compressed based on histogram using a sensitivity model of the human visual system (HVS). And for the detail layer, S curve is applied to depress the large deviations away from zero and enhance the low ones. The color balance correction method is adopted to suppress color distortion. Experiments and comparisons with state-of-art tone mapping algorithms demonstrate that our proposed algorithm not only effectively compresses the input image dynamic range, but also reserves the image details well. Simultaneously, the tone mapping images of our algorithm also keep the color, contrast and overall brightness perfectly.