基于目标红外特征与SIFT特征相结合的目标识别算法

A Target Recognition Method Based on Infrared Features and SIFT

为了解决传统的SIFT算法存在检测时间长,识别率低等问题,提出了一种基于目标红外特征与SIFT特征相结合的红外图像识别算法,该算法首先通过5个能反映红外目标初步信息且易实现的红外特征量进行初步识别,然后采用SIFT算法进行精确识别。通过三种飞机的红外图实验可以看出,将红外特征量与SIFT特征检测识别方法相结合,识别时间缩短0.06s,识别率有较大提高,达到98%以上。

In order to solve the problem of long detecting time and low recognition rate in the traditional SIFT algorithm, an infrared image recognition algorithm based on the target infrared features combined with the SIFT features is proposed. The algorithm first uses 5 infrared characteristic quantities that can reflect the infrared target preliminary information to realize the initial recognition, and then use the SIFT algorithm for accurate identification. Through three kinds of aircraft infrared experiment it can be seen that combining the infrared characteristic with the SIFT feature detection method, the recognition time is shorten to 0.06 s, the recognition rate is improved to more than 98%, and the proposed method has good application value.