自适应地空背景红外图像降噪增强方法

Research on the self-adaptive denoise and enhancement algorithm for infrared image under ground-sky background

针对地空背景红外图像对比度低,噪声干扰大的特点,提出一种自适应的地空背景红外图像的增强方法。通过基于场景的非均匀校正去掉大部分固定图案噪声,对降噪后的图像进行多宽度窗口联合的局部对比度增强。此算法只需代数运算,原理简单,效果显著,尤其适用于地空背景,海空背景或太空背景的红外图像目标探测与识别。此算法能够自适应地补偿红外图像固定背景噪声的漂移,使低对比度图像的各级细节得到增强,而且能够在压制背景噪声的同时使隐藏在暗处的人眼不可识别的细节清晰自然的显现出来。

Against to the characteristic of tow contrast and large noise for the infrared image under the ground-sky back-ground, a self-adaptive denose and enhancement algorithm is presented. The scene-based nonuniformity correction algorithm is used to remove the most fixed pattern noise. The locality contrast is enhanced based on the multiple windows united for the denoised image. The algorithm has a simple principle. All the calculations are algebraic, and have good results. It is very adaptive for the infrared target detection and recognition under the ground-sky background, the sea-sky background and the space back-ground. This algorithm can correct the drift of the fixed background noise of the infrared image, and enhance the details of the low contrast image.