A New Infrared Turbulent Fuzzy Image Restoration Algorithm Based on Gaussian Function Parameter Identification

Aiming at the image distortion caused by turbulence, an infrared image fuzzy measure algorithm based on local kurtosis of wavelet transform is proposed to identify the Two-dimensional (2D) Gaussian function model of the turbulence degraded image. In the process of image restoration, the Haar wavelet transform is applied to reduce the correlation of the image region. Then, the transformed coefficients are divided into sub-blocks and normalized. Combined with the characteristics of kurtosis, the local kurtosis of the whole image is calculated to describe the image ambiguity. The simulation results show that the proposed algorithm can effectively identify the parameters of the degradation model, and the relative error is about 5.5%. Compared with the existing algorithms, the algorithm can improve the rate of parameter identification and has higher deblurring efficiency.