Rock image weathering degree detection method and system based on deep learning

Abstract

The invention discloses a rock image weathering degree detection method and system based on deep learning. A neural network image recognition technology based on deep learning is applied to automaticdetection of the rock weathering degree, and the deep learning network is enabled to automatically output a rock weathering degree detection result by constructing and training the deep learning network. The automatic identification technology has relatively high accuracy; errors existing in traditional manual judgment of the rock weathering degree are reduced, the rock image recognition efficiency and accuracy are improved, and therefore the engineering labor cost is saved, real-time recognition and classification can be achieved, information can be input in time, the processing progress of field problems can be effectively tracked, image data are stored and shared on a platform, and an image classification management system is optimized.

Description

technical field [0001] The invention relates to the technical field of engineering geological survey, in particular to a method and system for detecting weathering degree of rock images based on deep learning. Background technique [0002] In the process of engineering geological survey, the determination and analysis of rock weathering degree is a basic work. The weathering degree of rock directly affects the judgment of engineering geological layering and foundation bearing capacity, which is an important reference for the engineering environment and material selection of the upper part of the rock. significance. According to the definition of rock hardness and weathering degree in GB50021-2001 "Code for Geotechnical Engineering (2009 Edition)" and TB 10077-2001 "Standards for Classification of Rock and Soil in Railway Engineering", the rock weathering degree is divided into five categories (unweathered, slightly weathered , weakly weathered, strongly weathered, fully wea...

AI Technical Summary

Problems solved by technology

[0003] However, the existing rock weathering degree detection work mainly relies on the artificial visual recognition method, which is greatly affected by subjective factors, and the artificial recognition detection results are not scientific, and are affected by on-site lighting, humidity and other conditions. There may be some differences in the judgment results

Therefore, in actual work, in order to reduce subjective errors, when judging the degree of rock weathering, a multi-person detection team is often formed, and repeated experiments and discussions are used to obtain more accurate detection and discrimination results, but this often increases the work of researchers. amount, and the identification efficiency is low, often resulting in waste of manpower and material resources

At the same time, there is a lack of an efficient recording method for the manual identification results, and researchers often use manual input recording methods, which will also lead to low efficiency in the detection of rock weathering and affect the quality of results

Images