基于Ohta颜色空间的石油地质图件目标分割算法研究

The Object Segmentation Algorithms Research in Petroleum Geological Drawing Based on Ohta Color Space

油田纸质地质图件矢量化意义重大。手工矢量化方法虽然能够达到精度要求,但劳动强度大,工作效率低。自动矢量化方法可提高工作效率,但出错率高,后期查错工作量大。针对自动矢量化方法的基础——目标分割问题,给出了基于颜色直方图和基于颜色表聚类的2个目标自动分割算法,可实现图件的高速分割处理。

Many geological drawings were accumulated in oil exploration and development history, vectorization of these drawings was great significance. To meet the accurate requirements, the manual vectorization is made, but the work is arduous, with low efficiency.Automatic vectorization is used to improve the work efficiency, but it has high error rate, therefore error-checking work is arduous. For the fundamental work of automatic vectorization-object segmentation, it is proposed that two-object segmentation algorithms are used in this paper, one is based on the color histogram the other is based on the color table cluster. Efficiency and robustness of these two algorithms are studied through a digital experiment.