最简边缘检测算子及其在地质建模中的应用

地质模型图的输入对于地球物理正反演的结果至关重要, 它将直接影响实际建模的质量。首次将图像边缘检测技术应用于地质建模领域, 实现了高效、高精度的地质模型输入。由于常规的边缘检测方法对于地质模型的边缘提取精度无法满足实际需要, 提出了一种最简边缘检测算子, 该算子是基于卷积核算子的最简形式, 最大检测误差仅为一个像素。实验结果表明, 将基于最简算子的图像边缘检测方法用于地质建模, 不仅保证了输入精度, 又大大提高了建模的效率, 很好地解决了地质建模中的精度和速度两大技术难题。

Simplest Operator Based Edge Detection Approach in Geological Modeling

Geological model maps play a primary role in the forward and inverse method of geophysics, and they will affect the results of geological modeling in practice. We apply edge detection to geological modeling for the first time, and achieve high efficiency and high precision. As far as geological model is concerned, many classical edge detection methods can't do well in precision that is the most important. A novel edge detection operator is proposed, which is the simplest operator based on regu-lar convolution kernel. Moreover, its error is not more than one pixel in the noise-free case. Experimental results prove that the presented method not only guarantees the high accuracy but also improves the efficiency greatly, that is to say, two technical problems have been solved by introducing image processing technique into geological modeling.