The application of the automatic search for visually similar geological layers in a borehole in introscopic camera recordings

• A method for query by image (QBIC) system in mining and geology was presented.
• Input material was introscopic camera recordings from exploring mining borehole.
• A method of searching by image visually similar geological layers was presented.
• The use of image processing, data analysis, classification is effective for QBIC.
• A set of image descriptors of geological layers in mining borehole were proposed.

The paper discusses the opportunities for obtaining valuable knowledge on the basis of video recordings from introscopic cameras, used to investigate mining boreholes. It also presents an original approach to the task of building a system, which would facilitate the process of searching for visually similar geological structures, and which would constitute an alternative to time-consuming manual profiling. The analysis of measurements performed with an introscopic camera is based on a feature space defined by the Authors, and uses the methods of mathematical morphology for processing video sequences, as well as data analysis for the decision-making process. As a result of the research, it was established that the methods of image processing and data analysis are an effective tool when it comes to automate the decision-making process with regards to the determination of the visual similarity of geological layers which make up a given, investigated area of a borehole. The methods in question yield a high percentage of correct decisions.