基于改进分水岭算法和Canny算子的医学图像分割

针对分水岭算法存在的过分割问题和医学图像的特点,比较并选取了彩色分量图像梯度信息的最大值,达到提取图像有效边缘信息的目的,然后利用阈值分割方法消除无效梯度信息,并用分水岭算法处理所得到的梯度图,再结合Canny算子提取的物体边缘得到分割结果.实验结果表明:该方法能够有效消除局部极小值和噪声干扰,得到精确的分割结果.

Medical Images Segmentation Based on Improved Watershed Algorithm and Canny Operator

With regard to the over-segmentation of watershed algorithm and the characteristics of the medical images,effectively edge information was picked-up firstly by comparing and choosing maxima from colour weight gradient images.Then,a method based on threshold segmentation was used for eliminating inefficacious gradient information.The new gradient image is segmented with watershed algorithm.Combining with the information of Canny edge detection,the segmentation result is obtained.Experimental results demonstrate that this new segmentation algorithm can not only effectively eliminate local minima and noise but also get exact segmentation results.