Application Of Improved Double Chain Quantum Genetic Algorithm In Image Segmentation And Denoising

Quantum genetic algorithm is a kind of efficient and parallel intelligent optimization algorithm based on the combination of quantum information and traditional genetic algorithm.Quantum genetic algorithm has the advantages of small population size,strong searching ability and fast convergence speed.But with the increasing demand of information technology for the speed and accuracy of large data processing,traditional optimization algorithm can not meet the needs in majority of occasions,thus prompting the scholars to improve it.Professor Li et al.put forward the double chain quantum genetic algorithm,which make up the defects of quantum genetic algorithm,improve the efficiency and precision,and get a wide range of applications.But there are still a lot of problems in the DCQGA,such as encoding space is too large to affect the search speed,and the rotation angle of the quantum rotation gate is not reasonable,which leads to the optimal value or the slow update and affects the speed of evolution.In this paper,the above problems were improved,and propose a new double chain quantum genetic algorithm which has higher search efficiency and stronger adaptive update step length B_DCQGA.Firstly,the solution space transformation method is improved to reduce the coding space,increase the search density and improve the search speed,under the premise of ensuring the single value mapping of coding space.Secondly,the adaptive step size factor is introduced in the quantum update,and the adaptive step size factor is optimized,so that the change of step size is more in line with the change trend of the optimal solution,which not only improves the problem of easily falling into local extremum caused by premature convergence phenomenon in most optimization algorithms,and overcome the update speed is too slow due to the low efficiency or update too fast over the optimal solution of the problems,which improves the searching speed and precision.Then the B_DCQGA algorithm is applied to the threshold selection mechanism of wavelet threshold denoising and the two-dimensional maximum entropy threshold segmentation.It can be proved through simulation,that B_DCQGA algorithm improves the denoising effect,increase the convergence speed and search accuracy of wavelet threshold function,can get smaller mean square error(MSE)and higher peak signal-to-noise ratio(PSNR),while retaining the most high frequency information,in the denoising of wavelet threshold.In the 2D maximum entropy threshold segmentation algorithm is introduced in B_DCQGA,through the original image and noisy image simulation experiments show that the B_DCQGA algorithm is effective to remove the noise,and reduce the evolution algebra,improve the efficiency of image segmentation.Through the application of B_DCQGA algorithm in two kinds of image processing,we can know that the algorithm has great advantage in the optimization ability,and can also be extended to more practical applications,and fully prove that the algorithm has a wide range of practicality and effectiveness.