In the process of acquisition and transmission

the infrared image of power equipment is easily affected by Gaussian and impulse noise

thus destroying the real information of the image. In order to meet the requirements of image quality to the process of image processing

a kind of filtering method aiming at the mixed noise of the infrared image is proposed in this paper. Firstly

the Gaussian noise in the infrared image is filtered out effectively by using trilateral weighted sparse coding (TWSC) model

and then the separation and removal of residual impulse noise is achieved by the combination of image structure and texture decomposition and median filtering algorithm so that the edge structure information of the image can be well preserved while the mixed noise of the infrared image is filtered. It is shown with combination of example that the method in this paper can effectively filter out noise and obtain higher peak signal to noise ratio (

PSNR

) as well as structural similarity (

SSIM

).