以计算机视觉的不变性理论为基础，用飞机红外图像边缘中的角点反映飞机的几何结构．然后对飞机的角点应用变化，以构造出两种不变特征，实验表明这两种特征具有一定不变性．推导了Ｍａｈａｌａｎｏｂｉｓ距离的平移、旋转和仿射不变性，并应用到空中目标的识别，实验表明这种方法效果良好，性能稳定 Based on the theory of invariance of computer vision, the aircraft’s geometry is reflected by the corners of the edge of the aircraft’s infrared image. Then the change of the aircraft corner is applied to construct two invariant features. Experiments show that the two features have some invariance. The translation, rotation and affine invariance of Mahalanobis distance are deduced and applied to the identification of aerial targets. Experiments show that this method has good effect and stable performance