基于改进的胡氏不变矩的坦克识别方法

Tank Recognizing Based on Improved Hu Invariant Moments

针对基于胡氏不变矩的坦克识别率低、在图像离散采样的情况下受图像尺度变化影响的问题,进行了图像尺度不变性研究,设计了一种基于改进的胡氏不变矩的坦克识别方法;为解决最大类间方差法在复杂背景下图像分割阈值与真实阈值发生偏移的问题,采用改进的最大类间方差法分割图像,减少了偏移量。将胡氏不变矩和改进的胡氏不变矩的算法进行了对比实验,实验结果表明:相对于采用胡氏不变矩的坦克识别方法,改进后的识别方法识别率提高了22%,该方法运算时间减少了80 ms,能克服图像的尺度变化造成的影响。

To solve the problem of low recognition rate of tank based on Hu invariant moments and the effect of the change of image scale in the case of discrete image, image scale in-variance was studied and a new method of tank recognition based on improved Hu invariant moments was designed; In order to solve the problem that the threshold of image segmentation threshold and the real threshold value are shifted in complex background, the improved maximum between-class variance was used to segment the image and reduced offset. A experiment was carried out to compare the invariant moments and the modified Hu invariant moments. The experimental results show that that compared with the method that Hu invariant moments is adopted to recognize a tank, and the recognition rate of the improved recognition increased by 22%, and the computation time of the method reduces 80 ms and it can overcome the influence of the scale change of image.