Robust adaptive enhancement algorithm for multi-modal high grayscale image displaying on low-bit monitors based on HIS and priori knowledge
Liangliang Li,Peng Wang,Jia Ren,Zhigang Lü,Ruohai Di,Xiaoyan Li,Hui Gao
DOI: https://doi.org/10.1016/j.displa.2023.102494
IF: 3.074
2023-07-14
Displays
Abstract:Gray-scale images are widely used in many fields, including satellite remote sensing, autonomous driving, medical CT, and X-ray detection, but gray-scale images have many shortcomings, such as high bit, low contrast, and high dynamic ranges. The visual analysis of these images is complicated and limited by people's psychology and physiology, and pseudo-color technology is an effective way to solve this problem. To address the weak adaptive capability, poor robustness, and insufficient generalization of existing pseudo-color enhancement methods, an adaptive enhancement algorithm for high grayscale images based on prior knowledge and HIS space, called APKHHIS, was proposed. Firstly, an adaptive gray-scale image correction mapping algorithm based on original prior knowledge was designed, and the quantization of images with different bit depths is realized. Secondly, to ensure the effectiveness of the enhanced data, an adaptive power compensation method was designed to compensate for the brightness of the gray-scale corrected images. Finally, an adaptive brightness and power correction algorithm was innovatively designed for the first time, which can adaptively adjust the color saturation and spatial brightness of the image more reasonably. To verify the effectiveness of the algorithm proposed in this paper, adaptive enhancement experiments were carried out on multi-modal data sets. The subjective and objective experimental results show that the APKHHIS algorithm can achieve the best performance on various data sets, and has strong robustness and universality, which provides an effective solution for the adaptive enhancement of multi-modal low-quality images.
engineering, electrical & electronic,instruments & instrumentation,optics,computer science, hardware & architecture