自动确定图像二值化最佳阈值的新方法

New Approaches for the Automatic Selection of the Optimal Threshold in Image Binarization

在详细分析二值化处理原理及当前技术的前提下 ,综合考虑图像的各种实际情况 ,提出了自动确定图像二值化阈值的增强大津法和边缘检测阈值法。多次实验表明 ,根据这两种方法确定的阈值对图像进行二值化处理 ,不仅可有效地解决图像灰度随地物特征变化的图像处理问题 ,同时对于同一信号源的图像也有相当理想的处理效果。

The selection of image binarization threshold is one of the critical technologies of the pretreatment of digital image. This paper proposes edge detection threshold methods of automatically selecting image binarization threshold after analyzing the principle of binarization operation and current technologies in detail. The result of experiments shows that using the threshold which is obtained from two methods to operate image binarization not only solves effectively image processing question of image gray's changing with object feature's changing but also can obtain ideal result of solving images getting from the same signal source.