面向快速成形的骨体断层图像分割算法研究

通过分析面向快速成形的骨体断层图像固有的特性，利用灰度特性对图像自动阈值分割．先对图像原始直方图进行移动平均修匀，再对其进行高斯函数的卷积运算以消除噪声点，通过有限次循环，获得平滑曲线。对卷积后的直方图作二次指数平滑处理，求得平滑信号的局部极值，即可确定最终的阈值．实验表明：该方法能快速、有效地分割骨体图像．

Segmentation algorithm for cross-section bone images in rapid prototyping manufacture

The properties of the gray to automatically segment the image with the acquired threshold was made use of based on the intrinsic properties of the bone in the image. To calculate the average of numerical values in the original histogram, the new function was convolutedly operated with gauss function to remove the noise in the image. By finite circulation operation the function curve got smoother and the part extreme value was obtained and so was the ultimate threshold. The experiments showed that the image can be segmented with great speed and high accuracy with the help of this algorithm.