Dam crack detection method based on unmanned aerial vehicle inspection and combined with deep learning

Abstract

The invention discloses a deep learning-combined dam crack detection method based on unmanned aerial vehicle inspection, and the method comprises the following steps: collecting dam surface images through an unmanned aerial vehicle, enabling two adjacent images to have an overlapping region with a set overlapping ratio, and completing the image collection of the whole dam through two collection modes of remote shooting and near shooting; preprocessing the acquired image; putting the pre-processed image and the original image into the constructed dam crack identification model for training to realize identification result fusion; segmenting crack pixels based on an image gray threshold, and completing crack segmentation according to connected domain analysis; and 5, after crack segmentation processing is completed on the original image, performing three-dimensional reconstruction by using oblique photography to obtain a dam three-dimensional model with crack information. According to the invention, rapid inspection and three-dimensional quantitative nondestructive testing of the structural crack can be realized, and the method has great engineering application value.

Description

technical field [0001] The invention relates to the field of structural health monitoring and damage identification of dams, in particular to a dam crack detection method based on drone inspection and combined with deep learning. Background technique [0002] Affected by factors such as disrepair and structural aging, the safety problems of large-volume concrete structures such as reservoir dams in my country are becoming more and more serious. However, the routine manual inspection of such structures is not only time-consuming and inefficient, but also cannot be inspected without dead ends. Therefore, it is becoming increasingly urgent to use advanced UAV intelligent inspection technology to conduct rapid daily inspections of large-volume concrete structures such as dams. [0003] At present, the rapid identification of structural cracks using image recognition technology with drones as the carrier has begun to enter the engineering application stage. However, the current...

AI Technical Summary

Problems solved by technology

However, the current technology is still in the two-dimensional stage and has major defects. For example, due to the interference of factors such as the shaking of the drone, the shooting distance, the focal length, external interference, and the complex background of the target, the image is often blurred and the area resolution is low. Problems such as low and too few crack pixels lead to low recognition accuracy of small targets, and it is difficult to accurately identify, locate, segment and quantify the damaged parts

Images