水下光场成像清晰度增强研究

Research on definition enhancement of underwater light field imaging

由于水介质及水下颗粒的散射吸收作用,使得水下成像图像存在分辨率与对比度低,细节特征模糊,颜色失真等问题。针对这些问题,该文提出基于光场成像的水下图像清晰度增强算法,根据暗原色先验理论和单尺度的Retinex理论建立水下散射成像模型对成像图像的清晰度进行增强处理,并利用光场成像的多视角特性对散射成像模型的图像增强效果进行优化,进一步提高水下成像图像的质量。实验结果表明,两种理论构建的水下散射模型和多视角优化算法可以有效的提高水下成像图像的质量。

Due to the scattering and absorption of water medium and underwater particles,there are some problems in underwater image,such as low resolution and contrast,fuzzy details,color distortion and so on.Aiming at these problems,this paper proposes an underwater image sharpness enhancement algorithm based on light field imaging.According to the dark priori theory and the single-scale Retinex theory,the underwater scattering imaging model is established to enhance the sharpness of the image,and the image enhancement effect of the scattering imaging model is optimized by using the multi angle characteristics of light field imaging,so as to further improve the quality of underwater imaging image.The experimental results show that the underwater scattering model and multi view optimization algorithm can effectively improve the quality of underwater image.