Study On Optical Characteristic Of Scattering Media

Point spread function is the core element of image processing, the basic parameter of image restoration and enhancement, and the major part of degenerate function. Applying MATLAB program, this thesis calculates the point spread function of the experimental images which possess different influence factors. Through the conclusion, comparision and summarization of the data, the author explores the effects of image color for point spread function and the influencing factors of image parameters. Thus it can lay a foundation for the construction and application of more extensive degradation model, and provide more accurate basis for the accuracy of image restoration.Motion sequence image has many applications in the three-dimensional image reconstruction, but it is a new attempt for the two-dimensional image restoration. Since non-uniform illumination has structural characteristic, motion sequence image can be considered as the continuous response distributed in space that optical properties of the target to the incident light. If the incident light distribution is regarded as the impulse response of a linear system, target optical characteristics extraction may represent the restoration problem of the non-uniform illumination system.Under a non-uniform, condition, how to recover target optical properties irrelevanting to lighting is the key issue to be solved in this experiment. This thesis presents a method of sequence images summation, and this summation constitutes the convolution illumination output of images unrelated to lighting and lighting distribution function. Summation operation eliminates the problem of non-uniform distribution of backscatter back ground resulting from non-uniform illumination and reduces the uncertainty of estimation from the point of statistics. And then we can apply image restoration methods to restore the image, obtain lighting unrelated images and verify the accuracy of the algorithm. So it can lay the theoretical foundation for the further restoration of underwater image and the estimation of three-dimensional shapes.