Event location identification method and apparatus

By segmenting the event stream and fusing features, the problem of low accuracy in visual position recognition in existing technologies has been solved, and higher accuracy visual position recognition has been achieved.

CN119741637BActive Publication Date: 2026-06-02SGCC GENERAL AVIATION +1
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SGCC GENERAL AVIATION
Filing Date
2024-12-12
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

Existing visual position recognition methods based on event cameras cannot provide a detailed description based on multi-dimensional information, resulting in low accuracy of visual position recognition.

Method used

By segmenting the event stream based on the timestamp of the event stream during the visual location recognition process, the segmented event stream fragments are represented in the form of voxel grids and converted into image frames. Temporal and spatial features are extracted, and feature enhancement algorithms and attention mechanisms are used for feature fusion. The spiking neural network is optimized to improve recognition accuracy.

Benefits of technology

It enables a multi-dimensional and refined description of event streams, improving the accuracy of visual location recognition.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119741637B_ABST
    Figure CN119741637B_ABST
Patent Text Reader

Abstract

The application discloses an event position recognition method and device, wherein the method comprises the following steps: cutting an event stream to obtain an event voxel; converting the event stream into an image frame; taking the event voxel as time information of the event stream and taking the image frame as space information of the event stream; extracting features of the time information and the space information to obtain time features and space features of the event stream; determining shared features according to the time features and the space features; performing feature enhancement processing and feature fusion on the time features, the space features and the shared features and giving different weights; migrating the corresponding weights to a pulse neural network to optimize the pulse neural network; extracting features of the event stream based on the optimized pulse neural network; matching the extracted features with features in an existing event position database; and determining position information of the event according to a matching result. The application can describe the event stream more finely and improve visual position recognition accuracy.
Need to check novelty before this filing date? Find Prior Art