The Study And Application Of Image Enhance And Image Segmentation Based On PCNN Optimized By GA

Pulse Coupled Neural Network(PCNN) is a new artificial neural network developed on basis of pulse synchronous oscillation in the visual cortex of small mammals in twentieth Century 90’s. With the characteristics of dynamic threshold,space-time summation,automatic wave propagation and synchronous pulse bursts, PCNN is emphasized in numerous fields such as image processing combinatorial optimization,artificial life and automatic target recognition.Due to the pulse coupled neural network on the basis of biology, can better simulate the biological visual neural system, it is suitable for image processing and its results is more in line with the human visual mechanism.This paper took the pulse coupled neural network in the application of image enhancement and image segmentation as the research subject. Some improved methods are proposed to deal with the complexity of PCNN structure and the deficiencies in image processing, as a result, the improved methods have a good effect in image denoising and image segmentation.The specific contents are as follows:(1)the structure and mechanism of pulse coupled neural networks were analysisted in-depth and its characteristics and application fields were introduced detailed.(2)The simplified PCNN model was introduced combined with its application in image processing. Because of the characteristics of image segmentation, the PCNN threshold function is improved,transforming the exponential decay function into linear attenuation function.(3)The activation mechanism of PCNN was applied to the detection of noise points to determine the location of noise in the image. Depending on the type of noise, we directly removed the noise in combination with other denoising method. This kind of method improved the efficiency of processing greatly.(4)Summarizing the concept of image segmentation and its common used methods. The foundation for PCNN being used to segment images are stated. A new image automatic segmentation algorithm is proposed with the combination of PCNN and the improved adaptive genetic algorithm. The algorithm makes PCNN parameters be set automatically and solved the problem of parameter setting difficulties. The experimental result shows that the algorithm has a good segmentation effect.(5)the new image segmentation algorithm is applied to cell division and project the split test of hot pepper particles,and have a good segmentation effect.The final paper put forward the suggestions and prospects for further research on the basis of summarizing the full text.