基于Guided Filter的地形图中线要素提取算法

针对彩色扫描地形图中线要素与背景要素难以分离，现有线要素提取算法提取的结果中存在边界不准确的问题，提出基于Guided Filter的地形图中线要素提取算法。利用基于能量密度和Shear变换相结合的线要素提取算法对线要素进行粗提取；引入Guided Filter，以源地形图图像作为Guided Filter的引导图像，以粗提取的线要素图像作为Guided Filter的输入图像，经过滤波处理获得的线要素信息更为显著；利用OTSU算法得到最终的线要素信息。实验结果表明：相对于现有的线要素提取算法，提出的算法能够更为准确地提取出地形图中的线要素，并具有更好的噪声抑制能力。

Line feature extraction algorithm from topographic maps based on Guided Filter

Aiming at problem that it is difficult to extract line features from topographic maps,and the line features extracted by exsting methods are not accurate,propose a line feature extration method based on Guided Filter. In this method,Guided Filter is introduced,the line features are extracted first by the method based on the energy density and the shear transformation,and the result images are used as the input images of Guided Filter;then use original topographic maps as the Guided images of Guided Filter. The line features are more significant after filtering,and obtain the final line features information by OTSU algorithm. Experimental resuts show that the proposed algorithm can extract line features more accurately,and has better ability of noise inhibition.