本文提出一种基于图像边缘几何特征的快速模板匹配算法。算法利用边缘

点的位置和梯度方向作为匹配信息进行相似度计算。可以很好的避免因图像明

暗变化、光照不均匀、旋转所带来的影响，且对于部分遮挡的情况，亦可以得

到良好的匹配结果。为了得到边缘点坐标和梯度方向，本文根据曲面拟合原理，

通过平移变换，推导出精确梯度方向和亚像素边缘坐标的快速算法。既加快了

算法的处理速度，也是匹配算法高精度的前提保证。为了使匹配算法满足实时

性要求，主要采用阈值判断和图像金字塔算法的搜索策略。在阈值判断方面，

通过选择一些特殊的边缘点作为试探点，加快了匹配度计算的效率，而且，很

好的避免了在部分遮挡情况下简单阈值判断方法的误判。最后，算法利用最小

二乘平差理论，通过一次近似处理，详细推导出快速获得亚像素级的定位和精

确的旋转精度方法-A fast template matching algorithm based on geometric features of image edges is presented in this paper. Algorithm using edge

The similarity calculation of the position and the gradient direction is used as the matching information. Can be very good to avoid because of the image

The effects of dark changes, illumination and rotation are not uniform.

To a good matching result. In order to obtain the coordinates of the edge points and the gradient direction, this paper is based on the surface fitting principle,

A fast algorithm for accurate gradient orientation and sub pixel edge coordinates is derived by using the translation transformation. Not only speed up

The processing speed of the algorithm is also the premise guarantee of the high precision of the matching algorithm. In order to satisfy the matching algorithm in real time

The search strategy of the threshold judgment and the image Pyramid algorithm is mainly used in the performance requirements. In the threshold judgment,