体表损伤图像的分割

图像分割的研究一直受到人们的高度重视，但是，针对体表损伤图像的分割还未见报道。基于利用体表损伤图像的特点，提出了两种适合损伤图像的分割方法。针对损伤部分和背景色彩相差剧烈的图像，提出了基于彩色图像分割技术路线：颜色调整、抽取RGB之一的分量和数学形态学处理。针对损伤部分的灰度比较均匀的图像，提出了基于活动轮廓模型（Snakes）的图像分割技术路线，共有7个步骤：图像增强、灰度化、中值滤波去除噪声、平滑边缘、提取初始轮廓、采用不同内外力参数的活动轮廓模型的迭代、B样条拟合。实验结果显示了这两种方法的有效性。分割结果为获得法医鉴定中的鉴定参数提供了准备工作。

Body-surface Trauma Image Segmentation

Image segmentation is highly concerned at all times, but body-surface trauma images segmentation is seldomly reported. In this paper,regard to the features of body-surface trauma images, two kinds of procedures are designed for the segmentation of body-surface trauma images. For the images that have acute contrast in colors between the trauma part and the background, a procedure is proposed that has four steps based on the image segmentation techniques: image enhancing, color adjusting, RGB heft extracting, mathematical morphology processing are implemented according to the set sequence. As for gray-average images,another method that has seven steps based on Snakes model is put forward. These steps in turn are image enhancing, image graying, median filter, edge smoothing, initial contour extraction, snakes model iteration with different parameters, and B sample imitation. Experiments with real body-surface trauma images are presented. The results show the effectiveness of this method. The results of segment images provide preparative work for the quantitative analysis of trauma parameters with rapidity and accuracy.