基于小波分析的二值化阈值选取

Binary threshold selection based on wavelet analysis

图像阈值的自动选取是灰度图像二值化处理中的关键。文章提出一种基于多尺度小波分析的阈值选取方法。通过对图像直方图特征点由粗到精的表示 ,在最佳尺度上确定出个数有限的波谷点 (即候选阈值点 ) ,再根据图像阈值的性质定出唯一的阈值点。该方法计算简单 ,适应性强 ,满足了实时性要求高的图像处理与分析系统。

Automation threshold selection is the key to binary processing of gray image. A method of threshold selection based on multi scale wavelet analysis is proposed. Because the characteristic points of imaging histogram can be represented by the characteristic of imaging threshold, the limited valley points can be attained in the best scale, and the unique threshold can be calculated according to the characteristic of imaging threshold finally. This simple method is especially effective for image processing and analyzing system.