A three-dimensional point cloud intrusion detection method and device based on deep learning

By using a deep learning-based 3D point cloud intrusion detection method and an EFT-RCNN network, the intrusion detection accuracy of coal storage yards has been improved. This solves the problems of misjudgment and latency in traditional methods and enables real-time, accurate intrusion prediction and hierarchical security response in complex environments.

CN121033758BActive Publication Date: 2026-06-19XIAN UNIV OF SCI & TECH
2 Cites 0 Cited by

Patent Information

Application Number
CN202511121708.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-08-12
Publication Date
2026-06-19
Estimated Expiration
2045-08-12

AI Technical Summary

Technical Problem

The coal storage yard environment is complex, and traditional detection methods are easily interfered with, leading to misjudgments and missed judgments. The existing lidar intrusion detection mechanism has a time delay and cannot provide sufficient security response time.

Method used

A deep learning-based 3D point cloud intrusion detection method is adopted. The EFT-RCNN network for personnel target detection is constructed using the Voxel R-CNN framework. Enhanced VFE, Focal Conv blocks and TeBEV Pooling techniques are used to improve feature representation and detection accuracy. Intrusion judgment is made by combining the 3D information of dangerous areas.

Benefits of technology

It significantly improves the intrusion detection accuracy at the coal storage yard loading site, realizes real-time and accurate intrusion prediction in complex environments, reduces the risk of missed detection and false detection, and provides a graded security response of prediction, early warning and alarm.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This invention discloses a method and apparatus for intrusion detection based on deep learning-based 3D point cloud. The method acquires a point cloud of the scene to be tested; it then performs target recognition on the point cloud using a personnel target detection network to obtain personnel target detection information. The personnel target detection network is constructed based on the Voxel R-CNN framework, and the feature set of the voxel grid is generated based on the position information of the points in the point cloud contained within the voxel grid and the point cloud features. Based on the detection information and the position information of dangerous areas in the scene to be tested, it determines whether an intrusion event has occurred. This invention compensates for the shortcomings of MeanVFE in feature utilization by reconstructing the voxel grid feature set generation method through the personnel target detection network, calculating the offset, and concatenating it with other features. It also achieves the aggregation of local and global information, utilizing feature information at different levels to improve feature expression quality, thus significantly improving detection accuracy.
Need to check novelty before this filing date? Find Prior Art