Cable appearance defect detecting and positioning method based on binocular vision

Abstract

The invention relates to a cable appearance defect detecting and positioning method based on binocular vision, which comprises the following steps that: a binocular camera respectively acquires defective cable images to obtain defective cable characteristics, forms a defective cable data set, performs target labeling on the data set by using a labeling box, and divides a training set and a test set; training the neural network by using the training set and the test set in the defective cable data set to obtain network weight parameters for detecting cable defects; network weight parameters used for detecting cable defects are deployed to edge equipment and are used for detecting the cable defects; performing target detection on a binocular cable image acquired by a binocular camera in real time by using the neural network model with the updated weight parameters; and solving a defect three-dimensional coordinate. The target distance is measured by adopting binocular measurement, and compared with an existing distance measurement method, the precision is higher. And the measurement range can be adjusted by flexibly adjusting the binocular baseline distance, so that the practical application is convenient.

Description

technical field [0001] The invention belongs to the technical field of cable appearance defect detection, and more particularly relates to an automatic detection method for cable appearance defect detection based on a neural network target detection algorithm. Background technique [0002] In recent years, national production activities have been inseparable from electricity, and social development has put forward new requirements for the safety and stability of electricity consumption. The stability of power transmission largely determines the experience of power consumption, and even affects production activities. Overhead lines are an early large-scale power transmission method, but they are easily affected by weather and environment and cause failures, occupy a large amount of land area, and easily cause electromagnetic interference to the surrounding environment. With the continuous development of the power system, this kind of The power transmission mode can no longer...

AI Technical Summary

Problems solved by technology

However, this method is not stable and is easily affected by the environment. Since most cables work in high-temperature and high-pressure environments, the temperature difference during cable operation is not easy to detect

[0008] To sum up, the current detection methods cannot meet the requirements of fast, efficient and automatic online detection of cable appearance defects at the same time, so a new method is urgently needed for the detection of cable appearance defects

Images