采用分区梯度模糊检测方法可以解决梯度边缘检测和模糊边缘检测方法在滤除背景时缺失前景要素的现象。利用提出的最优分量分色法,可以成功分离提取出前景要素;而在对提取出的线性要素进行细化时,提出了一种基于梯度特性的细化方法,得到的要素能很好地体现线性信息;采用现有的数学形态学方法对得到的细化要素进行后处理,最后得到了能够清晰表示要素特征的地图前景要素像素。
Based on the idea of gradient edge and fuzzy edge detecting, adopted fuzzy-gradient subarea method and it solved the problem which the general methods could not pick-up the fuzzy foreground elements from the map clearly. After removing the background of the map, this paper presented an optimal-component method to divide one kind foreground elements from the foreground and was successfully used in separating foreground. Then presented a method base on characteristics of gradient. Finally, it processed after-processing with mathematical morphology. The foreground elements getting from the map could refleet the characteristics of the map' s space and color. It was accurate and kept the form of the line very well. Experiments prove that running speed and the result are satisfied.