The Application of Adaptive Hybrid Filters in Infrared Image Denoising

Considering the characteristics of infrared images contaminated by Gaussian noise, impulse noise and other mixed noises, an adaptive hybrid filtering algorithm based on infrared image´s local statistical property is presented in this paper. It can automatically detect noises, determine the output based on threshhold value, select the size of filtering window, and judge the image edge on the basis of Laplacian Operator. The experiment results indicate that the presented algorithm has removed the infrared tank image´s mixed noises effectively, preserved the image details and edge features very well, and provided an effective way for infrared image denoising.