形态学图像处理下的矿石粒度的检测

Detection of Ore Granularity Based on Morphological Image Processing

利用工业摄像机对传送带上的矿石图像进行采集,将图像导入计算机利用matlab软件进行图像处理。对图像进行滤波处理、形态学运算、形态学重构、分水岭分割、矿石颗粒像素标定以及像素面积检测,然后统计矿石粒度的大小及不同粒级所占的比例,最后与人工筛分的矿石粒度作对比。实验表明通过图像处理得到的粒级比例与人工筛分的粒级比例两者差别较小,因此利用图像处理的方法能够较为真实反映矿石粒度的分布,提供矿石参数信息,准确性和实用性很高。

Using the industrial camera to collect the ore images on the conveyor belt,the image is imported into the computer and processed by matlab software.To filter the image,morphological operation,morphological reconstruction,a watershed segmentation,particle pixel calibration and pixel area detection,then statistical ore particle size and proportion of different grade,the final contrast and artificial sieving ore particles.Experiments showed that fraction ratio is obtained by image processing and artificial screening of graded ratio both difference is small,so the use of image processing method can more reflect the distribution of mineral particles,can provide a reference for field ore processing parameters,has strong practicability.