A method and system for real-time monitoring of water pollution based on UAV-borne multispectral imaging

By acquiring multi-source data using UAV-borne multispectral technology, and combining multi-source data fusion and spatial feature analysis, the problems of low cost, high coverage, and real-time performance in water pollution monitoring in existing technologies have been solved, enabling automated, precise, and intelligent early warning of water pollution.

CN122084850APending Publication Date: 2026-05-26JINAN GOLDENWORLD HIGHWAY INDUSTRY DEVELOPMENT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
JINAN GOLDENWORLD HIGHWAY INDUSTRY DEVELOPMENT CO LTD
Filing Date
2026-02-27
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing UAV remote sensing monitoring technology is difficult to achieve low-cost, high-coverage, and real-time water pollution monitoring. It also lacks the system fusion of multimodal data such as thermal infrared, making it difficult to distinguish between natural water temperature differences and industrial sewage thermal anomalies, resulting in a low level of intelligent early warning.

Method used

A real-time water pollution monitoring method based on UAV-borne multispectral imaging was adopted. By acquiring multi-source monitoring data, dynamic monitoring units were divided using a pre-defined classification model. Combined with a multi-source data fusion identification model and a spatial feature analysis model, abnormal pollution unit queues were determined and graded early warning information was generated.

Benefits of technology

It has achieved automation, precision, and integrated source tracing and early warning of pollution monitoring, thus improving the level of intelligence in water pollution monitoring and early warning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122084850A_ABST
    Figure CN122084850A_ABST
Patent Text Reader

Abstract

This application provides a method and system for real-time monitoring of water pollution based on unmanned aerial vehicle (UAV) multispectral imaging, belonging to the field of environmental water pollution monitoring technology. The method includes dynamically dividing multi-source monitoring data into monitoring units and generating pollution feature vectors corresponding to each dynamic monitoring unit; inputting the multi-source monitoring data and each pollution feature vector into a pre-trained multi-source data fusion and identification model to determine the abnormal pollution unit queue in the target water area; determining the source tracing parameters and pollutant occurrence state parameters corresponding to the abnormal pollution units based on a preset spatial feature analysis model; generating a pollution risk difference sequence for each abnormal pollution unit based on pollution risk values, source tracing parameters, pollutant occurrence state parameters, and a preset expert prediction model; and determining the early warning level based on the degree of difference of each pollution risk difference sequence to generate graded early warning information and send it to the ecological and environmental supervision terminal.
Need to check novelty before this filing date? Find Prior Art