并用正态核加权和密度函数拟合法解决了非正态图象模板匹配的优化问题。
By means of the density function fitting method of weighted sum of normal kernels the optimizing problem of template matching for non-normal distribution image is solved.
在指针式仪表自动判读系统中，首先需要对要判读的表盘图象与模板之间进行匹配，以确定所要判读的仪表类型与量程。
In automatic reading system of index-instrument, image of dial plate needed be read must match template in order to decide type and range of instrument needed be read.
提出图象的角点位于该矢势图的鞍点和基于模板匹配的鞍点检测算法，并给出实际的角点检测结果。
The saddle point detection algorithm is based on a simple template matching. The corner detection result is given.
应用推荐