IR image correction algorithm for turbulence-degraded

In order to deblur the fuzzy infrared image caused by high-speed turbulent flow field, a novel correction algorithm based on the modified incremental Wiener filter was proposed. Firstly, the degradation process was simplified as parameter-describing 2-D Gaussian function according to the prior knowledge, and the Point Spread Function(PSF) was estimated via an image quality assessment based blur identifier, which also can provide the initial restored image estimation. Then, strong edges in the initial restored image were detected and their corresponding area was smoothed to get the Edge Smoothed(ES) initial restored image. Finally, the ES image was restored through the modified incremental Wiener filter using the estimated PSF. Experimental results show that this algorithm can effectively suppress ringing artifacts, reduce the ringing metric of the restored image compared with those of the traditional iterative restoration algorithm and the Fuzzy filter based algorithm, and better the recovery image quality evidently, significantly reduce the time complexity.