Application of image enhancement based on Universal-FCMSPCNN

Abstract

In medical diagnosis, medical imaging technology is a potent clinical diagnosis approach. The focal examination and formation of diagnosis and treatment plans are greatly aided by the details and overall enhancement of medical images. It is critical to increase the amount of image information on the basis of high fidelity while studying image enhancement. This paper introduces a novel picture enhancement method and applies it to image processing using Pulse Coupled Neural Network (PCNN) research. Pulse coupled neural network is an artificial neural network that obtains temporal and spatial information from external stimuli and adjacent neurons. It has many unique excellent characteristics in various fields of image processing. Recently, we proposed an improved UFC-MSPCNN model based on the PCNN model. Firstly, we studied the PCNN model and MSPCNN model derived from PCNN model, and proposed this new model after analyzing its model principle and model complexity. Secondly, in our new algorithm, the synaptic weight matrix adopts a new setting method and redefines the attenuation factor α and the amplitude parameter V in the dynamic threshold. a new adjustment parameter J is defined to fine tune the dynamic threshold. Finally, we applied UFC-MSPCNN model to the image processing of left ventricular and peripheral lung cancer in the experiment. The experiment achieved good results, and the enhanced image accorded with the visual characteristics of human eyes, which proved the effectiveness of this method.