Application Of Genetic Algorithm In Optical Scanning Holography

Optical Scanning Holography(OSH)is a new kind of 3-D image reconstruction technology.The object's 3-D information could be stored in the two-dimensional hologram including image's amplitude and phase information.OSH can process image in real-time and has been widely used in scanning holographic microscope,3-D image recognition and remote sensing.In OSH,the traditional reconstruction method could be used to reconstruct the image from different layers.By redesigning the pupil,we can convert the defocus noise into speckle noise.In our proposed method,the two-dimensional maximum entropy method is combined with improved genetic algorithm(GA)to obtain the best threshold value of image segmentation and 2-D maximum entropy is used to distinguish the reconstructed image and separate it from the speckle-like noise.The improved genetic algorithm is to find the fitness value using overall searching ability.With these two methods combined together,we are looking forward to have a better reconstruction effect in the random phase encoded OSH system.This paper will focus on the application of genetic algorithm in optical scanning holography from the aspects of genetic algorithm theory analysis,MATLAB simulation and experimental verification.(1)Introduce the basic theory of OSH,fresnel diffraction theorem and mathematical model of the optical transfer function(OTF)in OSH system,including image information recording and reconstruction,then the defocus noise suppression methods.We conducted lots of OSH reconstruction image simulation research by means of MATLAB software.(2)Introduce the basic theory of genetic algorithm and the principle of two-dimensional maximum entropy method.Then we improve traditional genetic algorithm after analyzing the faults of traditional genetic algorithm.Comparing the simulation results in traditional genetic algorithm with which in improved genetic algorithm,we select the optimal improved GA method to get the best effect of image reconstruction.