基于自适应多尺度的血液细胞图像阈值分割方法研究

摘要

提出一个基于自适应多尺度滤波的阈值分割方法，对血液细胞图像进行了自动分割．首先，对图像原始直方图进行多尺度滤波，根据它的尺度空间图特性，确定合适的滤波尺度，然后完成对谷点由“粗”到“精”的定位，并作为阈值对原始图像进行分割；其次，利用松弛方法对分割图像作后处理；最后，将该方法与其他阈值分割方法做了分割性能的比较，证明了算法的有效性．%Based on an adaptive and scale space filter(ASSF),thresholding approach is proposed for automatic segmentation of blood cell images.Apply ASSF to image-histogram,a scale space image is built.According to the properties of the scale space image,a proper scale of filter is obtained.The valleys are gradually located from coarse level to fine level and the location of valleys are applied to the histogram-based segmentation.A relaxation-based postprocessing is employed for the segmented regions.The results show that the proposed algorithm is effective and robust.