A still weather satellite fire detection method fusing instantaneous and historical spatial differences

By constructing a historical background reference library and integrating instantaneous spatial differences, the problems of false alarms and missed alarms in existing fire detection methods have been solved, achieving higher accuracy and reliability in fire detection, and making it applicable to fire monitoring of multiple types of geostationary satellites.

CN122416609APending Publication Date: 2026-07-17QINGDAO STAR-RISING TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610274096.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-07
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing fire detection methods do not fully utilize time-dimensional information, ignore the daily variation of surface thermals, and are susceptible to interference from surface spatial heterogeneity and daytime heating effects, leading to false alarms and missed alarms. False alarms are particularly severe in highly heterogeneous areas, affecting the accuracy of fire monitoring.

Method used

A historical background reference library is constructed, and the deviations of the instantaneous spatial context and the historical background reference library are integrated. Fire point pixels are identified through multi-threshold discrimination conditions, and post-processing is performed to eliminate false alarms, thereby improving the accuracy and reliability of fire point detection.

Benefits of technology

It significantly reduces the false alarm rate and improves the detection capabilities of high-intensity fire points, low-temperature fire points, and fire points obscured by smoke. The algorithm maintains robust performance in different geographical areas and fire types, reducing false alarms and missed alarms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122416609A_ABST
    Figure CN122416609A_ABST
Patent Text Reader

Abstract

本发明涉及融合瞬时与历史空间差异的静止气象卫星火灾探测方法,该方法包括:首先利用多年的静止卫星观测数据,通过云掩膜、火点剔除及迭代统计滤波,构建无火晴空条件下的历史背景参考库。其次在火灾探测前,识别并剔除云、水像元,以去除潜在误报并确保后续背景统计值的准确性。在进行实时火点探测时,不仅计算目标像元与其周围背景的瞬时空间上下文差异,还引入该像元的历史统计特征,计算当前空间差异相对于其历史正常值的偏离程度。只有当瞬时差异和历史偏差同时满足预设的阈值条件时,才被判定为候选火点。最后通过后处理步骤剔除残余虚警,得到最终的火点探测结果。本发明通过融合瞬时与历史空间差异双特征作为判断依据,不仅有效抑制了由城市热岛、地物边界等固有热异质性及极端高温天气引起的误报,而且对低温火、早期火灾及烟雾遮挡火点具有高探测灵敏度,具备计算效率高、可移植性强的优点。
Need to check novelty before this filing date? Find Prior Art