基于小波变换的快速二维门限化方法复杂度研究

Research on the Complexity of a Wavelet Transform Based Rapid 2D Thresholding Method

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．

A two phased thresholding method is proposed which combines the wavelet transform (WT) with the traditional 2D histogram based thresholding method. The method obtains the range of a 2D threshold vector under the aid of WT in the first phase, and finds the accurate threshold vector from this range using traditional method in the second phase. The time complexity of the method is also discussed thoroughly. Theoretical analysis and experiment results both confirm the method’s rapidity and accuracy.