Crack image detection method based on Faster R-CNN parameter migration

Abstract

The invention discloses a crack image detection method based on Faster R-CNN parameter migration. The crack image detection method comprises the following detailed steps: 1) feature extraction: inputting a picture into a ResNet-50 network to extract features; 2) feature fusion and candidate region generation: inputting the obtained feature map into a multi-task enhanced RPN model, improving the size and the size of an anchor box of the RPN model to improve the detection and recognition precision, and generating a candidate region; and 3) detection processing: sending the feature map and the candidate region to a region of interest (ROI) pool, completely connecting the feature map and the candidate region to an (FC) layer, and then respectively connecting FC layer output to a boundary regression device and an SVM classifier to obtain the category and the position of the target. The crack image detection method solves the problem that dam crack image samples are insufficient, and is suitable for detecting cracks of the dam in different illumination environments and lengths.

Description

technical field [0001] The invention belongs to the field of image processing, in particular to a crack image detection method based on Faster R-CNN parameter migration. Background technique [0002] my country is the country with the most reservoirs and dams in the world, but affected by the natural environment and human factors, the surface and interior of the dam are deformed, and a series of dam appearance defects such as cracks, leakage, and calcification precipitation appear, which have a great impact The safety and durability of dam structures, cracks are one of the main hazards of dams. [0003] In recent years, the development of technologies such as image processing, pattern recognition and deep learning has provided technical support for dam crack image detection. Traditional machine learning methods require a large number of training samples and must satisfy the assumption that the training data and test data are identically distributed. Migration learning relax...

AI Technical Summary

Problems solved by technology

[0008] Purpose of the invention: In order to overcome the lack of images of dam cracks in the prior art, the use of smaller orders of magnitude samples for deep convolutional neural network training is prone to overfitting, and the Faster R-CNN algorithm has poor detection accuracy when dealing with multiple targets and small ta...

Images