基于图像处理的网格圈织物孔隙率检测

Inspection method of lattice apron porosity based on image processing

针对目前网格圈织物的孔隙测试相对困难的问题,采用图像处理方法对其孔隙及孔隙率的检测进行研究。首先对扫描所得的网格圈织物图像采用阈值分割方法将其转换成二值图像,对该二值图像利用投影法判断经纬纱直径和经纬密,计算得到理论孔隙率;然后对二值图像中纱线主体部分的假孔隙区域进行剔除,图像取反后再通过面积滤波得到仅包含真孔隙的网格圈织物二值图像,最后提取孔隙数量、孔隙面积尺寸、孔隙分布以及实际孔隙率等指标,并且对面积过小的孔隙进行位置标定。结果表明,用图像处理方法对网格圈织物进行检测,能够客观反映网格圈织物的孔隙率和孔隙分布的均匀性,且得到实际孔隙率远低于理论孔隙率的规律并解释了其成因。

It is rather difficult to measure the pore of lattice apron fabric nowadays. To solve this problem, the image processing method is proposed to assay the pores an porosity of the fabric. The scanned image of lattice apron fabric was converted into a binary image by threshold segmentation. The diameter of warp and weft yarns was determined by the projection data of the binary image. The theoretical porosity was calculated according the diameter and density of yarns. The interferentia] white regions on image were removed by judging whether they were in the body of the yarns or not. Then the binary image was reversed and filtered to remove small white regions. Pore characters such as the number of pores, area size of pores, the distribution of pores and actual porosity were extracted and small pores were marked. The experimental result shows that the pore characteristics by image processing method can objectively reflect the porosity of the lattice apron and the uniformity distribution of the pores. And the rule that the actual, porosity obtained is by far less than the theoretical valueand its causes are explained.