A novel approach for automated lithology extraction from core photographs

An automated lithology extraction methodology from core photographs using object based image analysis technology is developed. A knowledge based scheme (human interaction) is used for defining possible classes present in the image and for selection of appropriate samples (objects) that belong to each lithology classes. This can be saved separately and can be called using the program. Later a classifier (for automated classification) is trained to assign the unknown objects into appropriate classes based on the knowledge base. So the methodology can be quickly adapted to images from various geologic areas with adjustments in the knowledge base used to train the classifier. Hence this new methodology offers a fully automated lithology extraction methodology for any field. The automated classification calibrated to geologist s interpretation ensures the accuracy of the new methodology and the good match between the two gives the confidence to extend the use of this new methodology to the entire field. In addition, the methodology is applicable to several kinds of borehole images, for example wireline electrical borehole images and the more specialized LWD images.