云天背景下红外序列图像的小目标检测

DETECTION OF SMALL TARGET IN THE INFRARED IMAGE SEQUENCES IN CLOUD BACKGROUND

针对在云天背景下红外小目标的检测提出了一种背景抑制滤波器,与其它滤波器相比,滤波后图像的信噪比有了较大的提高。在图像分割时,先对图像均衡化,再利用最大对比度确定门限,该方法比其他分割方法具有更大的灵活性和有效性。在目标检测阶段,我们先进行HFQC滤波,然后设计管道作目标识别,能够较好地检测目标实验证明,该方法的检测精度有了很大提高。

A kind of background suppressing filter in cloud background for detecting small target in infrared image sequences is proposed. Compared with other filter, the higher SNR of image after filtering have been gained. During the course of image segmentation, the image equalization is done firstly. Then the image threshold is gotten by highest vision contrast ratio. The threshold is larger flexibility and practicality than the other image segmentation methods. In target detection, the sequences are filtered by the HFQC filterer, and then a pipe-model is designed for target recognizing. The results of experiment proof the method is very effective.