基于高斯型点扩展函数的红外图像热源复原

Heat source restoration of infrared images based on Gaussian point spread function

为了复原红外图像的热源,采用高斯点扩展函数的方法来增强热源的清晰度和对比度。首先,确定热源图像的高斯点扩展函数,建立其退化模型;然后,采用维纳滤波的方法复原红外图像中的热源,对复原图像通过YIQ变换来复原其温度场彩色信息;最后,通过边缘锐度和标准差评价分析热源复原的质量,并与盲复原算法对比。结果表明,边缘锐度边缘锐度和标准差分别提高了0.502%和0.124%。基于高斯型点扩展函数的方法对红外图像的热源复原具有明显的效果。

In order to restore the heat source in infrared images, Gaussian point spread function （PSF） was presented to improve the definition and contrast of infrared heat source. Firstly, the PSF of the heat source images was determined and the degradation model was established. Wiener filtering was used to restore the infrared radiation source and the color information of temperature field was restored by YIQ transform. Finally, the restoration image was evaluated quantitatively by edge acutance value （EAV） and standard deviation, and the restoration method was compared with blind restoration algorithm. The results show that EAV and standard deviation increase 0. 502% and 0. 124% respectively. The restoration method of heat source of infrared images based on Gaussian point spread function is effective obviously.