基于灰度选择法的医学图像可视化算法研究

针对医学融合图像可视化中存在病灶区域特征难以分辨这一问题,首先采用灰度阈值选择法实现对病灶区域的分割,然后在选定病灶区域的前提下,对融合图像做伪彩增强处理。实验采用的原始图像为可见光与红外源图像,融合是由采用小波变换图像融合法实现的,实现融合的区域为灰度图像。在医学图像中,由于肉眼对灰度图像不敏感,不方便清楚地观察病症部位,所以需要对病灶区添加明显的颜色特征。在对病灶区域的分割中采用了3种方法做对比,实验结果表明使用灰度阈值选择法得到的图像最为理想,由于灰度阈值选择法中阈值的选定比较繁琐,在该算法基础上结合最佳阈值分割迭代解法较准确地完成了阈值的选定。最后运用MATLAB实现仿真,实验表明该方法切实有效。

Medical image visualization algorithm based on gray level choice

There is problem that lesion area feature is difficult to distinguish in the medical image fusion visualization. Firstly, using the gray threshold selection method realize the segmentation of lesion area. Then, do false color enhancement processing to fusion image after selecting lesion area. The original images used in this paper are visible light and infrared images. The fusion of gray image is realized by using wavelet transform image fusion method. In the medical image, because of the naked eye is not sensitive to gray image and not convenient to observe lesion area clearly. So we need to add clear color features to lesion area. Three methods are used in the segmentation of lesion area. The experimental results show that image which obtained by using gray level threshold selection method is most ideal. It is complicated to selecting threshold value of gray level threshold selection method, so we combined optimum threshold segmentation iterative method to complete the selection of threshold value accurately. The simulation of MATLAB shows that this method is effective.