Kitchen hot work unattended alarm methods, systems, equipment and storage media

By using a dual-branch target detection network to perform real-time detection of kitchen video frames, the problems of false alarms and missed alarms in the kitchen hot work and unattended personnel alarm system have been solved, achieving accurate hot work and unattended personnel alarms and timely responses, and improving the level of intelligence in kitchen safety protection.

CN122135488APending Publication Date: 2026-06-02JIANGXI JISIDA KITCHEN PROD CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Applications(China)
Current Assignee / Owner
JIANGXI JISIDA KITCHEN PROD CO LTD
Filing Date
2026-01-29
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

Existing kitchen hot work and unattended personnel alarm systems have high false alarm and missed alarm rates, making it difficult to accurately determine whether a person is working while the fire is off-duty. They cannot achieve graded early warning and timely response, and cannot meet the refined and intelligent needs of kitchen safety protection.

Method used

A dual-branch target detection network is adopted, including a shared feature extraction module, a flame feature enhancement module, a human feature enhancement module, and a classifier. Through preprocessing and training of images of people and fire coexisting, flame images, and human images, real-time detection of kitchen video frames is achieved. If flames are continuously detected but no human is detected, the corresponding alarm or fire extinguishing response is activated.

Benefits of technology

It improves the accuracy and discrimination of flame and human detection, enhances the network's precision and computing speed, strengthens the ability to identify flame and human targets, and significantly improves the accuracy and robustness of alarms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122135488A_ABST
    Figure CN122135488A_ABST
Patent Text Reader

Abstract

This application discloses a method, system, device, and storage medium for alarming the absence of people in a kitchen during a hot work, belonging to the field of kitchen safety technology. The method includes: acquiring and preprocessing a dataset to obtain a target dataset; constructing a dual-branch target detection network, and training and testing the dual-branch target detection network based on the target dataset; acquiring target video frames of the kitchen in real time, and sequentially detecting and processing each target image in the target video frames according to the dual-branch target detection network to obtain the target detection result for each frame. This application, through shared feature extraction, branch feature enhancement, and independent classification decision-making of the dual-branch target detection network, maximizes the accuracy and discriminative power of flame and person detection while ensuring real-time processing speed, thereby making the monitoring results of each target image more reliable and improving the accuracy and computational speed of the network.
Need to check novelty before this filing date? Find Prior Art