基于LED辅助照明的水下图像增强算法研究

水下成像环境十分复杂,在成像过程中受到多种因素的影响,如光照的非均匀性,水对光的后向散射,成像器件的光学噪声,导致图像对比度低,细节模糊.针对水下图像的特点,提出运用同态滤波和双正交小波阈值滤波法相结合的图像增强算法改善水下图像光照的非均匀性、滤除噪声、提高水下图像的对比度,并通过峰值信噪比、均方差对该算法进行了性能评价.

Underwater image enhancement algorithms based on LED auxiliary lighting

Underwater imaging environment is very complex, in the imaging process is affected by a number of factors, such as the non-uniformity of the light, light backscattering, optical noise of the imaging devices, leading to low contrast image, detail fuzzy. According to the characteristics of underwater image, this paper proposes using the homomorphic filtering and the biorthogonal wavelet threshold filtering wears image enhancement algorithm improve underwater images illumination non-uniformity, filter out the noise and improve the contrast of the underwater images, and through the peak signal-to-noise ratio, mean square error of the proposed algorithm performance evaluation.