A structure icing / snowing synchronous classification monitoring system and method based on energy sensing and computer vision

By using a monitoring system that combines energy sensing multiplexing and computer vision with temperature difference and wind power generation, synchronous classification and intelligent early warning of icing/snow accumulation are achieved. This solves the problems of high cost, high power consumption and inaccurate monitoring of existing monitoring equipment, and is suitable for all-weather monitoring in frigid and snowy regions.

CN122368604APending Publication Date: 2026-07-10SOUTHEAST UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SOUTHEAST UNIV
Filing Date
2026-04-14
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing monitoring equipment suffers from high cost, high power consumption, single function, easy failure, and inaccurate monitoring in frigid and snowy regions, making it difficult to achieve unattended operation around the clock and simultaneous classification of icing/snow accumulation.

Method used

A monitoring system based on energy sensing reuse and computer vision is adopted, which combines temperature difference and wind power generation. Through composite energy acquisition and feature reverse perception, deep learning image feature fusion is used to achieve icing/snow accumulation classification, and load monitoring and data fusion are combined to achieve intelligent early warning.

Benefits of technology

It achieves low-cost, low-power, all-weather monitoring, improves monitoring accuracy, is suitable for distributed building clusters, has self-powered capabilities, and features intelligent early warning functions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure QLYQS_1
    Figure QLYQS_1
Patent Text Reader

Abstract

The application discloses a structure icing / snowing synchronous classification monitoring system based on energy sensing and computer vision, realizes composite self-power supply through a fan and a thermoelectric material, triggers a built-in high-definition camera of a deep residual network to perform morphological feature extraction in ice and snow weather, and realizes snow / icing / normal three-state classification identification for the first time; the output visual probability matrix, equivalent physical density and pre-physical prior feature are subjected to fusion analysis and judgment of multi-modal data, and all-weather autonomous monitoring and early warning of structure snow and icing synchronous classification are realized. Not only the problems of high energy consumption, difficult independent operation and single monitoring capacity of traditional equipment are effectively solved, but also precise synchronous classification of structure icing, snowing and normal state is realized, and the false alarm and missed alarm rates of traditional sensors are greatly reduced.
Need to check novelty before this filing date? Find Prior Art