基于HSV空间的岩心图像处理

Core Image Segmentation Algorithm Based on HSV Space

针对石油勘探中的岩心图像在分割过程中难以有效识别出灰度值相近的颗粒与颗粒、颗粒与背景的边缘信息,且容易造成过度分割或欠分割的问题,提出一种彩色图像分割并获取颗粒粒度分布的算法,并将结果与灰度迭代算法、最大熵法和分水岭算法分割结果进行比较。结果表明,该算法以彩色图像HSV模型的V分量为依据进行图像分割,可以在较为复杂的岩心背景下校正分割结果,更准确地分割出目标和背景颜色相近的边缘,获得砾石、泥沙颗粒的粒度分布。

An algorithm for color image segmentation and particles size distribution was proposed to solve the problem in oil exploration,which was core image segmentation being difficult to effectively identify the edge information between particles as well as particles and the background whose gray values were similar and likely to cause over-segmentation or less-segmentation.The results were compared with the segmentation results of gray iteration algorithm,maximum entropy method and watershed algorithm.The results show that the color image segmentation is based on the color information V component of the huesaturation-value(HSV) model.Correcting segmentation results in a more complicated background can be realized.The border of object and background with similar color can be accurately segmented.The particle distribution of gravel and silt particles can be obtained.