位场边缘识别特征点提取方法及应用

Extraction methods and application of feature points of edge recognition for potential field

位场边缘识别技术在地质构造特征研究中发挥着重要作用,但相关的自动化解释技术却发展缓慢.本文在位场边缘识别技术的基础上,研究了一套脊点提取、连接及特征线综合信息提取方法,可促进自动化地质解释的发展.首先,依据脊点的局部性质判断其存在性并计算其位置,采用分类连接、双向连接及排序等技术保证了脊点连接的连续性和高效性;之后,利用脊线方向、脊线笔直度、固定值线方向和固定值线扁度对特征线(脊线和固定值线)进行分区,并通过对特征线的方向、条数以及长度或面积进行综合统计,进一步提取边缘识别结果蕴含的信息.最后,通过实际资料测试验证了本文方法的有效性和实用性.

Edge recognition methods for potential field data play important roles in the study of geological structure features,but the related automated interpretation technologies have been developed slowly.In this paper,we studied a set of methods for the extraction and connection of ridge points of the edge recognition methods,to extract comprehensive information about feature lines,and to promote the development of automated geological interpretation.Firstly,the existence of ridge points is judged according to their local properties,and if they do exist,their locations are calculated then.After that,these ridge points are connected,and the continuity and high efficiency during the connecting process are ensured by several techniques such as classification connection,bidirectional connection and sorting.Next,the directions and straightness of the ridge lines,and the directions and flatness of the fixed value lines are used to partition the feature lines.Also,the directions,frequencies and lengths of the feature lines or the area circled by feature lines are counted for further comprehensive information extraction about the edge recognition results.Finally,the effectiveness and practicability of the proposed methods are verified by real data application.