POWER UTILITY TUNNEL SETTLEMENT CRACK IDENTIFICATION METHOD BASED ON ARTIFICIAL INTELLIGENCE TECHNOLOGY

WIPO Patent Application WO/2023/082418

A1

A power utility tunnel settlement crack identification method based on artificial intelligence technology. The method comprises the steps of (1) acquiring a tunnel wall image; (2) preprocessing the tunnel wall image, analyzing a situation of uneven illumination by using an image grayscale value, and performing histogram equalization correction processing; (3) performing image edge feature extraction on the preprocessed image; (4) performing BP neural network training by using a back propagation algorithm; and (5) inputting an extracted image edge feature into a trained BP neural network, and outputting a detection and identification result. Further provided is an improved back propagation algorithm suitable for tunnel wall crack detection in a complex environment. A large amount of image data collected in a power utility tunnel, including crack pictures and normal pictures of a wall of the internal structure of the tunnel, are used for learning and training, thereby finally realizing detection and identification of a tunnel settlement crack.

GU JIANJUN (CN)
WU LIUCHUANG (CN)
CHEN JIE (CN)
CAI RENLI (CN)
WAN HAO (CN)

PCT/CN2021/138598

May 19, 2023

December 16, 2021

Click for automatic bibliography generation  

STATE GRID JIANGSU ELECTRIC POWER CO LTD NANTONG POWER SUPPLY BRANCH (CN)
NANTONG POWER TRANS AND TRANSFORMATION ENGINEERING CO LTD (CN)

G06T7/00

CN111401131A	2020-07-10
CN110044268A	2019-07-23
CN112861681A	2021-05-28
CN109034530A	2018-12-18

NANTONG YONGTONG PATENT AGENCY (CN)

View/Download PDF      PDF Help

Previous Patent: GRABBING POINT INFORMATION OBTAINING METHOD AND APPARATUS, ELECTRONIC DEVICE, AND STORAGE MEDIUM

Next Patent: PLUNGER PUMP