2018/11/04 by Dana Berman, Berman, Dana, Deborah L. Levy +5 · 8 citations
Computer Science · Engineering · #Advanced Image Fusion Techniques #Advanced Image Processing Techniques #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Image Enhancement Techniques
paper · pdf · doi:10.48550/arxiv.1811.01343
openalex publication_date 2018/11/04 · openalex created_date 2022/08/02 · openalex updated_date 2026/07/28
Underwater images suffer from color distortion and low contrast, because\nlight is attenuated while it propagates through water. Attenuation under water\nvaries with wavelength, unlike terrestrial images where attenuation is assumed\nto be spectrally uniform. The attenuation depends both on the water body and\nthe 3D structure of the scene, making color restoration difficult.\n Unlike existing single underwater image enhancement techniques, our method\ntakes into account multiple spectral profiles of different water types. By\nestimating just two additional global parameters: the attenuation ratios of the\nblue-red and blue-green color channels, the problem is reduced to single image\ndehazing, where all color channels have the same attenuation coefficients.\nSince the water type is unknown, we evaluate different parameters out of an\nexisting library of water types. Each type leads to a different restored image\nand the best result is automatically chosen based on color distribution.\n We collected a dataset of images taken in different locations with varying\nwater properties, showing color charts in the scenes. Moreover, to obtain\nground truth, the 3D structure of the scene was calculated based on stereo\nimaging. This dataset enables a quantitative evaluation of restoration\nalgorithms on natural images and shows the advantage of our method.\n