论文部分内容阅读
本文提出了图象匹配中检测门限新的估计方法。研究了模板匹配的优化和最佳子模板实时估计的一般理论和方法。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。
In this paper, a new estimation method of detection threshold in image matching is proposed. The general theory and method of template matching optimization and real-time sub-template estimation are studied. The optimization problem of non-normal image template matching is solved by normal kernel weighting and density function fitting.