摘要
通过图像处理技术在带式输送机运输过程中对煤矸石的分割识别,以完成后续自动分拣是一项非常重要的工作。针对煤和矸石的光泽、灰度等差异,先通过灰度直方图确定阈值,将图像前背景相分离;再将图像距离变换,抽象为三维地形图,通过梯度下降算法把与矸石相接触的煤颗粒图像分离开。实验结果表明,该算法能有效地区分出煤与矸石。
Image processing techniques is a very important process, to complete subsequent automatic sorting coal gangue identification division in the belt transport. For coal and gangue gloss, gray and other differences, first through the histogram thresholdthe image before the background phase separation; then distance transform image, abstract three-dimensional topographic maps, gradient descent algorithm through thecontact with the waste rock coal particles are separated from the image.Experimental results show that the algorithm can effectively distinguish between coal and waste rock.
出处
《煤矿机械》
北大核心
2014年第8期128-130,共3页
Coal Mine Machinery
关键词
煤矸石
灰度直方图
距离变换
三维地形图
梯度下降算法
gangue
gray histogram
distance transform
dimensional topographic maps
gradient descent algorithm
