论文部分内容阅读
将小波分析这一新兴的数学工具与传统的基于灰度直方图的门限选择方法结合起来，提出了一种快速的二维门限化方法，并对其复杂度进行了深入地研究．该方法首先对二维直方图进行小波分解，得其低频分量，然后在此低频分量上确定出门限矢量的范围，最后在此范围内确定出精确的门限矢量．理论分析及实验结果均表明，该方法可以极大地降低二维门限化的时间复杂度．
Combining the emerging mathematical tool of wavelet analysis with the traditional threshold selection method based on gray histogram, a fast two-dimensional threshold method is proposed and its complexity is deeply studied. In this method, the two-dimensional histogram is first decomposed by wavelet to obtain the low-frequency component, and then the range of the threshold vector is determined on the low-frequency component. Finally, an accurate threshold vector is determined within this range. Theoretical analysis and experimental results show that this method can greatly reduce the time complexity of 2D threshold.