Method and apparatus for tracking small drone

The method uses Kalman filter and optical flow to track small drones in dynamic camera conditions, addressing tracking challenges by predicting object position and re-initializing upon failure, enhancing surveillance effectiveness.

KR102991608B1Active Publication Date: 2026-07-15ELECTRONICS & TELECOMM RES INST +1
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
ELECTRONICS & TELECOMM RES INST
Filing Date
2020-02-20
Publication Date
2026-07-15

AI Technical Summary

Technical Problem

Existing methods struggle to accurately track small unmanned aerial vehicles (drones) in dynamic camera environments due to rapid movement and lighting changes, making it difficult to maintain effective surveillance and prevent misuse.

Method used

A method combining Kalman filter-based target prediction and optical flow-based tracking to register feature points, predict object position and boundary area, and perform re-tracking upon failure, using a device with initialization, tracking determination, and measurement update units.

Benefits of technology

Enables efficient and accurate tracking of small drones in dynamic environments by detecting tracking failures and re-initializing, thereby improving surveillance capabilities and reducing tracking errors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure R1020200021025_ABST
    Figure R1020200021025_ABST
Patent Text Reader

Abstract

A method and device for tracking a small unmanned object are provided. The tracking device registers feature points of a region of interest in an input frame as points of interest, performs Kalman filter-based target prediction on the frame to obtain a prediction result, and performs optical flow-based tracking using the points of interest to obtain an estimation result. The tracking device then determines whether tracking of the unmanned object is successful by comparing the estimation result with the prediction result as a reference.
Need to check novelty before this filing date? Find Prior Art