针对文物图像的特征线提取方法研究

Feature line extraction method for heritage image

针对计算机辅助实现文物考古绘图自动化的问题,提出了一种文物图像的特征线提取方法。首先利用双边滤波器对文物图像进行预处理,改善了文物图像的主观质量;运用LoG边缘检测的方法得到图像的边缘路径;最后采用非均匀B样条得到插值边缘曲线,并利用高斯滤波器对其平滑处理来提高画笔生成的质量。实验结果表明,该方法是文物特征线提取的有效方法。

For the problem of automation of computer-aided cultural relics archaeology drawing, this paper presented a new method used to extract feature lines from a heritage image. First, it performed a bilateral filter to make a preprocess, aiming at improving subjective quality of the original image. Then it used LoG edge detection to get edge paths. At last, it got the interpolation edge curve using non-uniform B-spline, and made smoothing with Gaussian filter to improve the quality of result. Experiments show that it is an effective way to extract feature lines of a heritage image by this method.