Research On Bio-medical Image Restoration Algorithms

Digital image recovery technology plays an important role in the field of bio-medical image processing. A bio-medical image exists a certain degree of quality degradation and becomes a degraded image due to the noise, defocusing blur, motion blur. It brings great difficulties to the medical diagnosis and treatment. Therefore, it is necessary to study the restoration algorithms on the biological fluorescence image.At present, the research of image restoration algorithms has two main directions: one is to improve the accuracy of the restoration algorithms, make them restore the image details clearly; the other is to improve the efficiency of the algorithms, make them quickly restore the original image or give the best estimate of it in a relatively short time. According to the two targets, the former put forward some linear and non-linear image restoration algorithms. These algorithms meet some requirements of people to a certain extent, but still have their disadvantages: the linear algorithms are simple, but can not guarantee that the value of the image is non-negative in any space. It also can not restore lost frequency components; the non-linear algorithms are slow convergence, poor timeliness, and exist the problem of noise amplification. These problems have limited the applications of these algorithms.This paper focuses on the research of the Wiener filter restoration algorithm and Lucy-Richardson algorithm, and proposed a novel algorithm based on Lucy-Richardson algorithm. It introduces the basic theory of image restoration algorithm and several classical algorithms. According to the noise amplification of the Lucy-Richardson algorithm, wavelet analysis is introduced to the de-noising process. At the beginning of the algorithm, remove noise in images to reduce the influence of noise on Lucy-Richardson algorithm. Then according to the mathematical principle of Taylor series, vector extrapolation acceleration based on two orders Taylor Series is used to speed up the process of the iterative algorithm and reduce the number of iterations and computation time in each iteration of the algorithm. Then, verify the novel algorithm through the MATLAB platform. The experimental results show that the novel Lucy-Richardson algorithm proposed in this paper can effectively suppress noise interference in image, avoid the noise amplification to certain; and greatly accelerate the algorithm convergence speed, reduces the time complexity.