Dynamic target rapid sensing method and system based on deep learning

Abstract

The invention discloses a dynamic target rapid sensing method and system based on deep learning. The method comprises the following steps: carrying out environment sensing based on sensing system joint calibration, and collecting data from different data sources; based on the collected data, scene acquisition is carried out, and map data and pose information are acquired; performing normalized data preprocessing of multivariate features based on the obtained map data and pose information, and performing feature matching on the data after the normalized data preprocessing through superpixel segmentation and semantic analysis; and based on data obtained by feature matching, sensing the dynamic target by using a sparse convolutional network and motion estimation fusion. According to the invention, the technical problem of low dynamic target identification accuracy in the prior art is solved.

Description

technical field [0001] The present invention relates to the field of AI intelligence, in particular to a method and system for fast perception of dynamic targets based on deep learning. Background technique [0002] With the development of power technology and the improvement of power quality requirements of power users, the dynamic target of substations needs to quickly perceive the technical construction needs urgently. The current state of the art is as follows: [0003] (1) The current reduction of construction land leads to the reduction of equipment spacing. Due to the lack of precise sensing and positioning technology and methods, the construction difficulty of live work near the substation is increased, and the safe operation space is reduced; [0004] (2) The working space of the urban line channel becomes crowded, and there is a lack of active dynamic identification and judgment of non-contact safety distance; [0005] (3) The sensing sensitivity and distance acc...

AI Technical Summary

Problems solved by technology

[0003] (1) The current reduction of construction land leads to the reduction of equipment spacing. Due to the lack of precise sensing and positioning technology and methods, the construction difficulty of live work near the substation is increased, and the safe operation space is reduced;

[0004] (2) The working space of the urban line channel becomes crowded, and there is a lack of active dynamic identification and judgment of non-contact safety distance;

[0005] (3) The sensing sensitivity and distance accuracy of the existing distance sensing devices based on field strength theory on the market are greatly affected by voltage, electromagnetic field, and transmission environment medium

[0006] To sum up, due to the current unfavorable target positioning and identification supervision, the safety production situation of substations is severe, with high incidence of personal casualties and equipment damage accidents

Although a variety of detection algorithms have been developed for dynamic object perception detection, most object perception methods based on deep learning rely on a large number of annotation labels, require a lot of work, and are not suitable for and uncooperative spatial objects. And the versatility of these algorithms needs to be further improved, and the accuracy and robustness of the algorithms are not strong enough to quickly and accurately lock the target and calculate the state and category parameters of the dynamic target in a complex environment

In addition, the extraction of different feature values ​​often only focuses on the information of a certain aspect of the image, and cannot express the target more comprehensively.

[0007] For the above problems, no effective solution has been proposed

Images