Full text loading...
-
Template Matching Based Automated Detection of Curves from Scanned Raster Log Images
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 78th EAGE Conference and Exhibition 2016, May 2016, Volume 2016, p.1 - 5
Abstract
Well log curve records the physical, electrical or other properties of the rock/ fluid mixtures penetrated by drilling a borehole into subsurface. These logs are then converted into raster images which are further digitized for geological interpretation. The digitization process is usually manual and time consuming. Merely few researchers have proposed the automated algorithms in order to perform effective and efficient digitization. However, this paper proposes a quick, robust and fully automated method to detect and digitize curves from scanned raster log images. The proposed algorithm is novel and is based on template matching and adaptive filtering techniques to extract the curves from raster images. The extracted curves are then digitized by picking the bright spots from the candidate image. Total 87 raster log images containing single and multiple curves are automatically digitized in this research with the overall accuracy of 91.52%. Additionally, this paper also presents the detailed comparison of the proposed algorithm with the commonly used radon transform where our proposed method gave better results in terms of accuracy.