In image dehazing, due to the influence of sky area and local light spot, dehazing images often have problems with brightness and light spot. Therefore, accordinng to the prior theory of dark channel, an image dehazing algorithm based on region segmentation and block optimization is proposed. Firstly, the quadrant tree cycle method is used to find the region with the highest local brightness in the whole image, and the atmospheric light value is estimated through this region. Secondly, combined with the image information, the image is divided into a bright region, transition region and dark region, and the transmittance of each region is compensated and controlled. Finally, the dehazing image is obtained by using the atmospheric scattering model, and the brightness of the image scene is improved by singular value decomposition. The experimental results show that the restored image is bright and natural, with obvious details and high color fidelity. The algorithm effectively solves the defect that the dark channel prior algorithm fails in the sky area and highlight area, and can restore a more natural sky highlight scene.
Image Dehazing Algorithm Based on Region Segmentation and Block Optimization
2022-10-12
1577120 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Underwater image and video dehazing with pure haze region segmentation
British Library Online Contents | 2018
|Springer Verlag | 2021
|Image dehazing based on haziness analysis
British Library Online Contents | 2014
|Prior based Single Image Dehazing using Decision Image
IEEE | 2020
|Single image dehazing based on vector quantization
British Library Online Contents | 2015
|