图像识别算法在油田集输仿真培训中的应用

Application of Image Recognition Algorithm in the Training System of Gathering Simulation

以石油企业中的油气集输系统为主要研究对象,针对目前虚拟仿真技术存在的问题和缺陷,对油气集输系统的虚拟仿真技术进行较深入研究.在对油气集输工具图像进行分析和实验研究的基础上,提取集输工具图像的多种特征值,其中利用HSV彩色空间提取图像颜色特征,采用傅立叶描述子提取图像的形状特征.并采用最小二乘支持向量机算法对图像进行识别,实现了员工在虚拟现实场景中真实、安全、高效培训,经现场实际使用取得了较好的应用效果.

This paper deeply researches the virtual simulation technology of oil gas gathering transferring system in the view of the present problem and defects existing in the virtual simulation technology. It treats oil gas gathering and transferring system in oil companies as main research object. The system extracts multiple eigenvalues of gathering and transferring tools＇ images which is on the basis of analysis experiment research in the image of oil gas gathering and transferring tools. This algorithm uses the HSV color space to extract images＇ color features and takes advantage of fourier descriptor to shape features. It identifies images by least squares SVM algorithm which finally implements real, safe, etlicient training of staff, and has achieved good application affect after actual using.