一种基于热成像原理的图像热噪声去除算法

IR Image Denoising Algorithm Based on Principle of Thermal Imaging

针对气动热环境下高温头罩对红外成像系统造成的热辐射噪声干扰问题,提出了一种基于热成像原理的图像热噪声去除算法。首先,利用有限取样光线代表连续光线,在头罩内对取样光线进行反向追迹,分析得到头罩温度和传感器热辐射噪声成像灰度之间的关系;其次,利用Vega软件仿真得到一定大气条件下点光源通过高温头罩的红外图像;最后,根据前面得到的头罩温度和热噪声成像灰度之间的关系,计算得到高温头罩的红外成像灰度分布,将热辐射噪声视为加性噪声,从目标的红外图像中减去该噪声得到去除热噪声后图像。实验结果表明,该方法去除热辐射噪声效果明显,图像质量得到了改善,弥补了硬件制冷的不足,为下一步的图像复原打下了很好的基础。

An infrared image denoising algorithm based on the theory of thermal imaging is put forward to suppress the thermal radiation noise in infrared imaging system caused by high temperature optical window in the aero-dynamic environment. Firstly, limited sample ray is used to represent the continuous ray which is traced reversely, the relationship between the temperature of the optical window and the image gray scale of the thermal noise is calculated. Secondly, the infrared image of a point source which is through the high temperature optical window under some atmospheric condition is simulated by the Vega. Finally, the infrared imaging gray level distribution of the optical window is calculated based on the relationship between the temperature and the image gray scale. The thermal radiation noise is regarded as additive noise which is removed from the infrared image of the target. The results of the experiment show that the method to remove thermal radiation noise takes effect obviously and the infrared image quality is therefore improved. The method makes up for the deficiency of the refrigeration hardware and lays a good foundation for the next step of image restoration.