Power transmission channel fire sparse time sequence detection method, system, device and storage medium

CN121861594BActive Publication Date: 2026-05-29HUAYAN INTELLIGENT TECH (GRP) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HUAYAN INTELLIGENT TECH (GRP) CO LTD
Filing Date
2026-03-19
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

In existing methods for detecting wildfires along power transmission lines, target detection based on single images is prone to misidentification, while video-based identification suffers from huge storage space requirements and excessive time consumption, making it difficult to meet the timeliness requirements for wildfire detection.

Method used

A sparse temporal detection method is adopted, which detects wildfires by single images and combines a multi-round voting mechanism of temporal images. It uses texture, color and shape change features to make progressive judgments, thereby improving the accuracy and timeliness of wildfire detection.

Benefits of technology

It significantly improves the accuracy of wildfire detection, reduces the false alarm rate, avoids high computational and storage overhead, and balances the requirements of detection accuracy and real-time performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121861594B_ABST
    Figure CN121861594B_ABST
Patent Text Reader

Abstract

The application discloses a power transmission channel mountain fire sparse time sequence detection method and system, equipment and a storage medium, and relates to the technical field of mountain fire detection. The method comprises the following steps: inputting a single power transmission channel image into a mountain fire suspicion detection model to obtain a mountain fire suspicion detection result; if there is a mountain fire rectangle frame, a group of candidate suspicion images are collected based on a preset adjacent time interval; the group of candidate suspicion images are input into the mountain fire suspicion detection model to obtain a group of candidate suspicion detection results; the group of candidate suspicion images are respectively cropped according to the group of candidate suspicion detection results to obtain a group of candidate suspicion ROI images; and multi-round voting is performed according to the group of candidate suspicion ROI images to obtain actual mountain fire detection results. In this way, the accuracy and authenticity of mountain fire identification are improved, the false positive rate is significantly reduced, high calculation and storage overheads caused by continuous video analysis are avoided, and the detection accuracy and real-time requirements are balanced.
Need to check novelty before this filing date? Find Prior Art