Study Of Image Restoration Method Based On Improved Particle Swarm Optimization Algorithm

In the process of image generation, storage and transmission, often because of hardware equipment, weather conditions, light intensity, noise and a number of factors such as image quality degradation. In order to recover these degraded images, a number of mature image restoration methods, including inverse filter image restoration method, Wiener filter, image restoration, to recover from the noise, geometric distortion correction method. Inverse filtering is simple, but the effect is unsatisfactory handling noise. Wiener filter combination of degradation and noise statistical properties of the function of two aspects, but need to know the Wiener filter noise without degradation of image and power spectrum, these two parameters more difficult to obtain. To recover from the noise and geometric distortion correction can be used by noise pollution, the image or the image geometrical distortion serious. The above algorithms have some limitations.In many application fields, intelligent algorithms can solve optimization problems efficiently. Genetic algorithm is an evolutionary intelligent evolutionary algorithm, it by evolving the initial solution and gradually close to the optimal solution, which includes selection, crossover and mutation operation, the evolutionary process is to spend on behalf of the variation of the next generation, to the value of fitness function whether the threshold for the sign to the purpose of optimization. The PSO is a swarm intelligence algorithm, it has the function of global optimization, through the exchange of information between groups to solve optimization. But the swarm intelligence algorithm iterative process in offspring selection mechanism is not clear, although the iteration speed is faster than the evolution of intelligent algorithms, but convergence is not as good as evolutionary algorithms, easy to fall into local optimum.In the use of swarm intelligence algorithm for image restoration process, this thesis introduced the selection mechanism of genetic algorithm into the particle swarm algorithm and proposed a new image restoration techniques which is based on improved particle swarm algorithm . IPSO algorithm in order to analyze the process of image restoration application effect, this article using the IPSO algorithm with the traditional genetic algorithm, the standard particle swarm algorithm were degraded image restoration. Experimental data show that the standard genetic algorithm for image restoration effect in the ideal, but the computing efficiency could not be satisfactory; standard particle swarm optimization parameters less operation, high efficiency, but the recovery as effective as genetic algorithm. The improved particle swarm algorithm for IPSO combination of genetic algorithms and particle swarm optimization advantages in the recovery effect basically meet the requirement, but also in the computing efficiency is higher than the standard genetic algorithm.A new idea of this thesis is introduce selection process of genetic algorithm into standard particle swarm optimization, and IPSO algorithm is applied to improve image restoration. Experimental results show that the IPSO algorithm in different situations during the rehabilitation of degraded images showed good recovery of computing efficiency and effectiveness in the application of image restoration has good practical value.