标题:
岩石铸体薄片图像孔隙自动提取方法The Automatic Extraction Method of the Pore of the Rock Casting Body Image
作者:
代贺, 滕奇志, 伦增珉
关键字:
岩石铸体薄片, H分量, 样本训练, 自动提取Rock Casting Body, H Components, Sample Training, Automatic Extraction
期刊名称:
《Journal of Image and Signal Processing》, Vol.6 No.1, 2016-12-22
摘要:
针对传统岩石铸体薄片图像孔隙提取耗时、效率低的问题，提出了一种自动快速提取方法。首先，由于采用染色树脂或液态胶加工生成的铸体薄片其特征区域均有一定的颜色，例如：深蓝、浅蓝、深红、浅红等，通过对铸体薄片特征区域颜色的分析发现不同颜色的铸体薄片其彩色空间H(色调)分量分布具有一定的集中性；其次，采用样本训练，即计算出样本特征区域的H分量的阈值；最后，利用该阈值对同批铸体薄片图像进行直接提取即可。这样就实现了岩石铸体薄片图像的孔隙自动提取。
To solve the problems of time consuming and low efficiency in the traditional extraction of pore analysis from rock casting body image, an entirely new method of extraction of pore analysis from rock casting body image was proposed. Firstly, the rock casting body, which is produced by the pigmented resin or liquid glue, has a certain color, such as dark blue, light blue, dark red, light red and so on. The different characteristic regions have a certain concentration of H components by the analysis of the characteristic region. Secondly, using sample training, that is to say, the threshold value of the H components of the sample characteristic region is calculated. Finally, by using the threshold value of the H components, the same batch of rock casting body image can be extracted directly. In this way, we can realize the automatic extraction of the pore of the rock casting body image.