基于多尺度小波变换的同名点匹配方法研究

针对一类特殊小波变换系数不同尺度下的模极大点能提供在不同尺度下信号急速变化点的位置信息,提出了一种基于小波变换的模极大位置的图像点为匹配基元的同名点匹配方法。该方法采用由粗到细的匹配策略,应用顺序约束,减少了小尺度的小波变换模极大值的位置的图像点的匹配空间,为了避免误匹配和误差传播,舍弃了以小波变换的模极大值比较作为匹配准则,采用了基于米字条形窗口的互相关函数的匹配准则。这种特征匹配与区域匹配相结合的方法,解决了同名点准确匹配难、计算量大等问题。采用该方法对篦冷机内水泥熟料高度进行测量,实验表明该方法能较精确的得出水泥熟料料层的分布状况。

Research of correspondence points matching based on multi-scale wavelet transformation

A method based on the image point of local maxima position of a wavelet transformation is presented to solve correspondence points.It is based on that the multi-scale local maxima of wavelet transform can obtain the locations of the signal sharp variation points at scales for a particular class of wavelets.A coarse to fine matching algorithm is applied to solve it.The algorithm decrease matching space of the point in the position of local maxima of a wavelet transformation of smaller scale using sequence restriction.In order to avoid mistaken matching and error transmission,we applied the matching rule of cross correlation function basing on the strip window instead of the matching rule of local maxima of a wavelet transformation comparison.Thus the complexity of correspondence points matching is decreased and the matching mistake is reduced using the area-based matching method coupled with feature-based method.The thickness of cement clinker in the cooler is measured using this method.The experiment indicates that the distributing status of cement clinker thickness can be calculated accurately.