红外目标湍流退化图像的优化复原算法

提出了基于最速下降法的湍流退化图像盲目复原算法。将图像转换到频域中，建立一个基于目标图像和点扩展函数频谱的目标函数，通过迭代方式采用最速下降优化方法来极小化该目标函数，并利用傅里叶变换和反变换将目标图像和点扩展函数在频域和空域之间进行变换，在每次迭代中交替加入约束条件进行反复修正，以便取得预期的图像恢复效果，增强算法的稳定性和抗噪能力。针对红外目标湍流退化图像，在微机上对算法进行了一系列复原验证实验。实验结果表明：该文算法复原效果稳定，抗噪能力强，具有实用价值。

Optimization restoration algorithm for infrared object turbulence-degraded image

A blind restoration algorithm for turbulence-degraded image based on the steepest descent method is proposed.An objective function based on frequency spectrums of the object image and the point-spread function(PSF) is set up in the frequency domain,which is minimized by the steepest descent method in an iterative manner.FFT and IFFT are used to transfer the object image and the PSF between frequency domain and time domain,and the constraints of the frequency domain and space domain are introduced in each iteration to modify them,so as to obtain the expected image restoration effect,the proposed algorithm robustness and better immunity to noise.A series of restoration experiments for infrared object turbulence-degraded images are performed to test the proposed algorithm in the microcomputer,and the experimental results show that the proposed algorithm is robust and immune to noise.