红外双色亚图像识别研究

The Research of Infrared Bicolor Quasi Image Recognition

研究了红外双色辐射特性和双色比识别特征 ,分析了红外亚成像的灰度与空域特征。研究了双色比区分伪目标 ,灰度扩展亚成像方法。研究了亚图像子块灰度直方图的迭代二值化 ,二值分割后的空域模糊分割。研究了单元扫描亚图像识别的特征参数 ,初始成像多帧配准识别实现低虚、漏警率目标检测 ,并通过锁定目标实施连续逐帧分割、识别。实验表明 ,所研究的红外双色亚成像识别方法适用于静止。

Infrared radiation and bicolor ratio are studied in the paper. The gray and space feature of infrared quasi imaging is analyzed. Bicolor ration discernibility is applied to reject false targets. Gray expansion is used in quasi imaging. Iterative sub-block gray histogram bilevel is studied for quasi images, and then fuzzy spatial segmentation is adopted for 2-bit image. Recognition feature of unit scan imaging is presented. Infrared quasi image frames registration is used to improve detection rate. The recognition method minishes the original false and omission possibility. Successive segmentation and recognition with original locking are studied. The experiment reveals that infrared bicolor recognition is proper for stationary, slow, and fast infrared targets and it can be realized in real time.\;