Research Of Image Edge Detection Method Based On Pulse Coupled Neural Network

In the process of digital image processing,it is often used to process thousands of image information data,the workload is very large.If the representative information of the image can be extracted,the efficiency of image processing will be improved significantly.The edge is an important feature of the image which concludes the outline and structure information of different objects.People often segment the image into different goals and areas based on the edge feature,then understand and identify these object,good edge detection results can greatly reduce the data of subsequent image processing while preserved image feature information.Therefore,how to extract the edge of image quickly and accurately has become a research hotspot in the field of image processing.The common edge detection methods are simple in design and fast detection.But they have some disadvantages: inaccurate edge location,poor continuity and the sensitive to noise.On the basis of the edge characteristics of digital image,this paper proposes an image edge detection algorithm based on pulse coupled neural network(PCNN)which can detects and extracts the image edge effectively.As an important symbol of the third generation artificial neural network,the PCNN neuron is very similar to the human visual neuron.But it has a large number of network parameters,each parameter may affect the performance of the entire network.So the parameter setting becomes more complex and critical.This paper introduces the ignition characteristics of the PCNN model in the case of non-coupling connection and coupling connection,then improves the network parameters of simplified PCNN model and optimizes the key parameters adaptive setting method.Before processing,the image may be contaminated by various kinds of noise pollution.These noise pollution not only hinder the visual effect of the image,but also affect the results of image processing seriously.This paper analyzes the characteristics of the Salt and pepper noise and Gaussian noise.The noise point are detected and extracted by using the synchronous pulse release characteristics of PCNN.Then the noise points are filtered by mathematical operation.This method can reduces the blurring of image texture and detail in the image edge detection.This paper designs an improved PCNN model according to the phenomenon of edge detection in digital image PCNN iterative processing.The feasibility and performance of the algorithm is verified by Matlab simulation and compared with the traditional algorithm.How to determine the adaptive setting method of the key parameters based on the characteristic of digital image is the key to improve the performance of this algorithm,the paper carries out research and analysis in detail for this.