Method and system for monitoring position of fire extinguisher in hotel environment based on monitoring video

Abstract

The embodiment of the invention discloses a method and system for monitoring the position of a fire extinguisher in a hotel environment based on a monitoring video, and the method comprises the steps: comprehensively collecting video monitoring captured by each camera in a hotel, obtaining a real-time video stream, extracting features, inputting the features into a feature extraction model, judging whether the position information of the fire extinguisher accords with preset position information or not, and if yes, carrying out the operation; if yes, current position information is recorded, detection is continued, and if not, an alarm is sent, and the moving track of the fire extinguisher is tracked. The embodiment of the invention is used for improving the automation level and accuracy of fire extinguisher position monitoring in public places, especially in internal environments of hotels, and ensuring the effectiveness and accessibility of fire extinguishing equipment, so that the overall fire safety management level is improved; the accuracy, the automation degree, the response speed, the environmental adaptability, the cost control, the flexibility and the like are remarkably improved, and powerful support is provided for safety management of public places.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of image recognition technology, and in particular to a method and system for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video. Background Art

[0002] In the current field of fire safety management, especially for the location monitoring of fire extinguishers in large public facilities such as hotels, it mainly relies on traditional manual inspection methods and some preliminary automation solutions. These existing technologies mainly include the following methods:

[0003] Manual regular inspection: This is the most common practice, which involves arranging special personnel to regularly inspect and record the status and location of fire extinguishers. Although this method is direct and effective, it has significant disadvantages, such as high labor costs, low efficiency, and prone to omissions or incorrect records. Especially in large facilities, due to the wide area, the inspection work is time-consuming and laborious, and it is difficult to ensure timeliness and accuracy.

[0004] Sensor-based positioning systems: Some places try to use RFID tags or other types of sensors to track the location of fire extinguishers. However, the deployment cost of such systems is high and each fire extinguisher needs to be specially modified and installed. In addition, in complex environments (such as those with metal obstacles), signal transmission may be unstable, affecting positioning accuracy.

[0005] Video surveillance systems: Traditional video surveillance is mainly used for post-event review and does not have the ability to automatically identify the location of fire extinguishers. Although some advanced systems integrate motion detection functions, they are not effective for monitoring static objects such as fire extinguishers and cannot adapt to the challenges brought by environmental changes (such as changes in lighting conditions and occlusion).

[0006] Simple image recognition based on computer vision: In recent years, studies have proposed using computer vision technology to identify specific objects, including fire extinguishers. However, early methods are usually based on hand-crafted feature extraction algorithms with limited robustness and accuracy, especially in the face of complex backgrounds and changing viewpoints. Summary of the invention

[0007] To this end, the embodiments of the present invention provide a method and system for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video, so as to solve the technical problems that the prior art is expensive, difficult to implement and has poor adaptability, lacks intelligent analysis capabilities, is difficult to ensure the real-time and accuracy of data, and cannot cope with complex scenarios in actual applications.

[0008] In order to achieve the above purpose, the embodiment of the present invention provides the following technical solutions:

[0009] According to a first aspect of an embodiment of the present invention, a method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video is provided, the method being applied to a high-definition camera network and comprising:

[0010] S1. Comprehensively collect video surveillance captured by various high-definition cameras in the hotel, obtain real-time video streams and slice the real-time video streams to obtain video slices;

[0011] S2. Construct a convolutional neural network model, train the convolutional neural network model using the fire extinguisher feature data, and generate a trained feature extraction model;

[0012] S3, extracting features from the video slices and inputting them into the feature extraction model, comparing them with the fire extinguisher feature data, and generating a comparison result;

[0013] S4. Use the comparison result to determine whether the current position information of the fire extinguisher conforms to the preset position information. If so, record the current position information and continue the detection. If not, send an alarm and track the movement trajectory of the fire extinguisher.

[0014] Furthermore, the video surveillance captured by each high-definition camera in the hotel is fully collected, the real-time video stream is obtained and the real-time video stream is sliced ​​to obtain video slices, including:

[0015] Slice the real-time video stream by frame to obtain video frame slices;

[0016] The data enhancement algorithm is used to perform data enhancement on the video frame slices to obtain video slices.

[0017] Furthermore, features are extracted from the video slices and input into the feature extraction model, and compared with the fire extinguisher feature data to generate comparison results, including:

[0018] extracting historical fire extinguisher feature data at the current location from a fire extinguisher feature database;

[0019] The current fire extinguisher feature data extracted from the video slice is obtained and compared with the historical fire extinguisher feature data at the current location to generate a comparison result.

[0020] Furthermore, the fire extinguisher feature data is a multi-scale fire extinguisher feature obtained by fusing fire extinguisher features of different scales.

[0021] Furthermore, the multi-scale fire extinguisher features obtained by fusing the fire extinguisher features of different scales include:

[0022] Acquire original image data and extract low-resolution features, medium-resolution features and high-resolution features from the original image data;

[0023] Fusing the low-resolution features, medium-resolution features and high-resolution features to form a comprehensive feature vector;

[0024] The comprehensive feature vector is the multi-scale fire extinguisher feature.

[0025] Furthermore, if the conditions do not match, an alarm is sent and the movement trajectory of the fire extinguisher is tracked, including:

[0026] Compare the current frame and historical data in real time and record the timestamp, and generate the position change data of the fire extinguisher through the comparison results;

[0027] The position change data and the corresponding timestamp are recorded in the log.

[0028] Furthermore, if the condition does not meet the requirement, an alarm is sent and the movement trajectory of the fire extinguisher is tracked, which also includes:

[0029] Generate an alarm notification by using the fire extinguisher's movement information, location change data and corresponding timestamps;

[0030] The alarm notification is sent to a security management terminal.

[0031] Furthermore, the sending form includes SMS, email or application push.

[0032] Furthermore, the method further comprises:

[0033] The current time and light information in the video stream are obtained, and the current time and light information are combined to perform light compensation on the video frame slices in the video stream.

[0034] According to a second aspect of an embodiment of the present invention, a fire extinguisher position monitoring system in a hotel environment based on surveillance video is provided, the system comprising:

[0035] The acquisition module is used to comprehensively acquire video surveillance captured by various high-definition cameras in the hotel, obtain real-time video streams and slice the real-time video streams to obtain video slices;

[0036] A training module, used to construct a convolutional neural network model, train the convolutional neural network model using fire extinguisher feature data, and generate a trained feature extraction model;

[0037] A comparison module, used for extracting features from the video slices and inputting them into the feature extraction model, comparing them with the fire extinguisher feature data, and generating a comparison result;

[0038] The detection module is used to use the comparison result to determine whether the current fire extinguisher position information conforms to the preset position information. If it conforms, the current position information is recorded and the detection is continued. If it does not conform, an alarm is sent and the movement trajectory of the fire extinguisher is tracked.

[0039] The embodiments of the present invention have the following advantages:

[0040] The method described in the embodiment of the present invention is applied to a high-definition camera network, which includes: S1, comprehensively collecting video surveillance captured by each high-definition camera in the hotel, obtaining real-time video streams and slicing the real-time video streams to obtain video slices; S2, constructing a convolutional neural network model, using fire extinguisher feature data to train the convolutional neural network model, and generating a trained feature extraction model; S3, extracting features from the video slices and inputting them into the feature extraction model, comparing them with the fire extinguisher feature data, and generating a comparison result; S4, using the comparison result to determine whether the current fire extinguisher position information meets the preset position information, if it meets, recording the current position information and continuing to detect, if it does not meet, sending an alarm and tracking the movement trajectory of the fire extinguisher. The embodiment of the present invention is used to improve the automation level and accuracy of fire extinguisher position monitoring in public places, especially in the internal environment of hotels, ensure the effectiveness and accessibility of fire extinguishing equipment, and thus improve the overall fire safety management level, and achieve significant improvements in accuracy, automation, response speed, environmental adaptability, cost control, and system flexibility, effectively solving many problems existing in the prior art, and providing strong support for the safety management of public places. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0042] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0043] Figure 1 A schematic diagram of the logical structure of a fire extinguisher position monitoring system in a hotel environment based on surveillance video provided by an embodiment of the present invention;

[0044] Figure 2 A flow chart of a method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video provided by an embodiment of the present invention;

[0045] Figure 3 A schematic diagram of the architecture from image acquisition to final alarm in a method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video provided by an embodiment of the present invention;

[0046] Figure 4 A schematic diagram of a feature extraction process in a method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video provided by an embodiment of the present invention;

[0047] Figure 5 A schematic diagram of position identification logic in a method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video provided by an embodiment of the present invention;

[0048] Figure 6 A schematic diagram of an abnormal alarm mechanism in a method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video provided by an embodiment of the present invention;

[0049] Figure 7 A schematic diagram of multi-scale feature fusion in a method for monitoring the location of a fire extinguisher in a hotel environment based on surveillance video is provided as another embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] In the current field of fire safety management, especially for the location monitoring of fire extinguishers in large public facilities such as hotels, it mainly relies on traditional manual inspection methods and some preliminary automation solutions. These existing technologies mainly include the following methods:

[0052] Manual regular inspection: This is the most common practice, which involves arranging special personnel to regularly inspect and record the status and location of fire extinguishers. Although this method is direct and effective, it has significant disadvantages, such as high labor costs, low efficiency, and prone to omissions or incorrect records. Especially in large facilities, due to the wide area, the inspection work is time-consuming and laborious, and it is difficult to ensure timeliness and accuracy.

[0053] Sensor-based positioning systems: Some places try to use RFID tags or other types of sensors to track the location of fire extinguishers. However, the deployment cost of such systems is high and each fire extinguisher needs to be specially modified and installed. In addition, in complex environments (such as those with metal obstacles), signal transmission may be unstable, affecting positioning accuracy.

[0054] Video surveillance systems: Traditional video surveillance is mainly used for post-event review and does not have the ability to automatically identify the location of fire extinguishers. Although some advanced systems integrate motion detection functions, they are not effective for monitoring static objects such as fire extinguishers and cannot adapt to the challenges brought by environmental changes (such as changes in lighting conditions and occlusion).

[0055] Simple image recognition based on computer vision: In recent years, studies have proposed using computer vision technology to identify specific objects, including fire extinguishers. However, early methods are usually based on hand-crafted feature extraction algorithms with limited robustness and accuracy, especially in the face of complex backgrounds and changing viewpoints.

[0056] Problems and shortcomings of existing technologies:

[0057] 1. Manual inspection is inefficient, costly, and difficult to ensure the real-time and accuracy of data.

[0058] 2. Sensor-based solutions are expensive, difficult to implement, and limited by the physical environment.

[0059] 3. Traditional video surveillance lacks intelligent analysis capabilities and cannot proactively detect changes in the location of fire extinguishers.

[0060] 4. Early computer vision solutions relied on fixed rules, had poor adaptability, and were unable to cope with complex scenarios in practical applications.

[0061] In summary, the current fire extinguisher location monitoring methods all have problems to varying degrees, and a more efficient, accurate and easy-to-implement technical solution is urgently needed. The present invention aims to overcome the above-mentioned shortcomings of the prior art and improve the safety management level of public places by combining deep learning and video analysis technology.

[0062] In order to solve the above-mentioned technical problems such as high cost, difficulty in implementation and poor adaptability, lack of intelligent analysis capabilities, difficulty in ensuring the real-time and accuracy of data, and inability to cope with complex scenarios in actual applications.

[0063] refer to Figure 1 The embodiment of the present invention discloses a fire extinguisher position monitoring system in a hotel environment based on surveillance video, the system comprising: an acquisition module 1; a training module 2; a comparison module 3; and a detection module 4.

[0064] Use advanced image processing technology and pattern recognition algorithms to automatically detect and locate specific objects (such as fire extinguishers). By monitoring the position changes of fire extinguishers in real time, potential safety hazards can be discovered and corrected in a timely manner to ensure the safety of personnel and property. Combine video surveillance systems with IoT technology to support remote monitoring and management functions, and enhance the system's response speed and flexibility. Use deep learning models for feature extraction and classification to adapt to complex environmental conditions and changing target appearances.

[0065] The present invention aims to solve the problem of low efficiency and error-proneness in the existing technology of fire-fighting facilities such as fire extinguishers being inspected manually, and to provide an efficient, accurate and easy-to-implement technical solution suitable for all kinds of public and commercial places that require a high level of fire safety protection. Through innovation in this technical field, not only the intelligent level of fire extinguisher management is improved, but also a theoretical basis and technical support are provided for further exploration of other similar application scenarios.

[0066] Corresponding to the above-disclosed fire extinguisher position monitoring system in a hotel environment based on surveillance video, the embodiment of the present invention further discloses a fire extinguisher position monitoring method in a hotel environment based on surveillance video. The following describes in detail a fire extinguisher position monitoring method in a hotel environment based on surveillance video disclosed in the embodiment of the present invention in combination with the above-described fire extinguisher position monitoring system in a hotel environment based on surveillance video.

[0067] refer to Figure 2 The present invention discloses a method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video, which is applied to a high-definition camera network, and comprises: S1, comprehensively collecting video surveillance captured by each high-definition camera in the hotel, obtaining a real-time video stream and slicing the real-time video stream to obtain video slices.

[0068] HD camera network deployment: Install HD cameras in key areas of the hotel (such as corridors, halls, outside rooms, etc.) to ensure full coverage of locations where fire extinguishers may be placed.

[0069] Multi-angle coverage strategy: By rationally arranging the angles and positions of cameras, it ensures that there are no blind spots in the target area.

[0070] S2. Construct a convolutional neural network model, train the convolutional neural network model using fire extinguisher feature data, and generate a trained feature extraction model.

[0071] Deep learning model training: A convolutional neural network (CNN) is used to train a feature extraction model that is specifically optimized for the appearance features of fire extinguishers.

[0072] Application of data enhancement technology: In order to improve the robustness of the model, data enhancement technology (such as rotation, scaling, brightness adjustment, etc.) is used to simulate image changes under different lighting conditions and viewing angles.

[0073] S3. Extract features from the video slices and input them into the feature extraction model, compare them with the fire extinguisher feature data, and generate comparison results.

[0074] Real-time feature matching: The features extracted from the real-time video stream are compared with the pre-trained fire extinguisher feature database to determine the specific location of the fire extinguisher.

[0075] Multi-scale feature fusion: Combine feature information at different scales to improve recognition accuracy, especially in the presence of occlusion or partially restricted field of view.

[0076] See also Figure 4 , using deep learning models (such as CNN) to extract unique visual features of fire extinguishers from video streams and compare these features with a pre-trained database to determine their location.

[0077] S4. Use the comparison result to determine whether the current position information of the fire extinguisher conforms to the preset position information. If so, record the current position information and continue the detection. If not, send an alarm and track the movement trajectory of the fire extinguisher.

[0078] Dynamic tracking mechanism: Once the fire extinguisher is detected to be moved or lost, the system can automatically track its trajectory and record the timestamp and location information of the relevant events.

[0079] Instant alarm function: When an abnormal situation is detected, the system will immediately send an alarm notification to the security management personnel, including SMS, email or application push message.

[0080] Furthermore, the video surveillance captured by each high-definition camera in the hotel is comprehensively collected to obtain a real-time video stream and slice the real-time video stream to obtain video slices, including: slicing the real-time video stream by frame to obtain video frame slices; and performing data enhancement on the video frame slices using a data enhancement algorithm to obtain video slices.

[0081] Furthermore, features are extracted from the video slices and input into the feature extraction model to be compared with the fire extinguisher feature data to generate a comparison result, including: extracting historical fire extinguisher feature data at the current location from a fire extinguisher feature database; obtaining the current fire extinguisher feature data extracted from the video slice and comparing it with the historical fire extinguisher feature data at the current location to generate a comparison result.

[0082] Furthermore, the fire extinguisher feature data is a multi-scale fire extinguisher feature obtained by fusing fire extinguisher features of different scales.

[0083] In the case of occlusion or limited field of view, it is necessary to improve recognition accuracy by combining feature information at different resolutions. Figure 7 , a multi-scale fire extinguisher feature is obtained by fusing fire extinguisher features of different scales, specifically comprising: acquiring original image data and extracting low-resolution features, medium-resolution features and high-resolution features from the original image data; fusing the low-resolution features, medium-resolution features and high-resolution features to form a comprehensive feature vector; the comprehensive feature vector is the multi-scale fire extinguisher feature.

[0084] Environmental adaptability adjustment: Automatically adjust image processing parameters according to different environmental conditions (such as day / night mode switching) to ensure high recognition rate in various situations.

[0085] Online learning capability: The system has online learning capabilities, which can continuously update and improve the model based on newly collected data to continuously improve performance.

[0086] refer to Figure 5 ,The location of the fire extinguisher is determined by comparing the feature vector extracted in real time with the template in the feature database. If it matches, the location record is updated; if it does not match, the anomaly detection mechanism is activated.

[0087] Further, refer to Figure 5 If it does not match, an alarm is sent and the movement trajectory of the fire extinguisher is tracked, including: real-time comparison of the current frame and historical data and recording the timestamp, generating the position change data of the fire extinguisher through the comparison result; recording the position change data and the corresponding timestamp in the log.

[0088] Figure 6 When a fire extinguisher is detected to be moved or lost, the system automatically generates an alarm message and notifies the security management personnel through various methods such as SMS, email or application push.

[0089] Further, refer to Figure 6 If it does not meet the requirements, an alarm is sent and the movement trajectory of the fire extinguisher is tracked, which also includes: generating an alarm notification by using the fire extinguisher movement information, location change data and corresponding timestamp; and sending the alarm notification to the security management terminal.

[0090] Furthermore, the sending form includes SMS, email or application push.

[0091] Furthermore, the method also includes: acquiring current time and light information in the video stream, and combining the current time with the light information to perform light compensation on video frame slices in the video stream.

[0092] refer to Figure 3 , Figure 3 The entire system architecture from image acquisition to final alarm is shown. The HD camera network is responsible for capturing environmental images. After image acquisition, feature extraction, and location recognition, the status monitoring and abnormal alarm module processes and notifies relevant personnel.

[0093] The present invention aims to solve the problems of inaccurate, inefficient and manual monitoring of fire extinguisher locations in existing hotels and other complex environments. The specific goals are as follows: Improve monitoring accuracy: By adopting deep learning and video analysis technology, fire extinguishers in hotel environments can be automatically and accurately identified and located to reduce false alarms and missed alarms. Enhance automation level: Real-time monitoring of the location of fire extinguishers can be achieved, and the location information of fire extinguishers can be automatically recorded and updated without human intervention, significantly improving management efficiency. Optimize response speed: When a fire extinguisher is detected to be moved or lost, the system can quickly issue an alarm to ensure that security managers can take immediate action to reduce potential risks. Adapt to complex environments: Advanced image processing algorithms are used to overcome challenges such as lighting changes and occlusions to ensure that the system can operate stably in various complex environments. Reduce costs: Through automated monitoring, the demand for human resources is reduced, while avoiding the high deployment cost of traditional sensor positioning systems, providing an economical and efficient solution to improve the safety of public places and ensure the availability and accessibility of fire extinguishing equipment in emergency situations. It not only improves the level of fire safety management, but also provides new ideas and technical support for similar application scenarios.

[0094] The embodiment of the present invention realizes the full process automation from image acquisition to final alarm, greatly reduces the need for manual intervention, and achieves a high degree of automation; it adopts advanced deep learning algorithms, which can accurately identify fire extinguishers in complex environments and determine their specific locations; it has efficient anomaly detection and alarm functions, which can detect potential safety hazards and notify relevant personnel in the first time; it supports multiple alarm methods and is easy to integrate into the existing fire management system, with good compatibility and scalability.

[0095] The embodiments of the present invention have the following advantages:

[0096] 1) Improve monitoring accuracy

[0097] Deep learning algorithms are used for feature extraction and recognition, combined with data enhancement technology to simulate different environmental conditions. Compared with traditional rule-based image processing methods, deep learning models can automatically learn the unique features of fire extinguishers from a large number of samples and improve recognition accuracy by continuously optimizing model parameters. This significantly improves the accuracy of fire extinguisher location recognition in complex backgrounds and reduces false positives and false negatives.

[0098] 2) Enhance the level of automation

[0099] By building an integrated video analysis model, the entire process from image acquisition to location recognition to abnormal alarm is automated. The high-definition camera network is used for real-time monitoring, and the information in the video stream is automatically analyzed through the background algorithm. The location of the fire extinguisher can be updated and recorded without human intervention, which greatly improves management efficiency, reduces labor costs, and ensures the timeliness and accuracy of data.

[0100] 3) Optimize response speed

[0101] Implement a dynamic tracking mechanism. Once a fire extinguisher is detected to be moved or lost, an alarm is triggered immediately and the event details are recorded. The current frame can be compared with historical data in real time to quickly detect location changes and send alarm notifications through multiple communication channels (SMS, email, etc.), which greatly shortens the response time and enables security managers to take action at the first opportunity to reduce potential risks.

[0102] 4) Adapting to complex environments

[0103] The use of multi-scale feature fusion technology and online learning capabilities enables the system to maintain high recognition rates under different lighting conditions and in the presence of occlusions. Multi-scale feature fusion technology can comprehensively consider information at different levels of detail, while the online learning function allows the system to self-adjust and optimize based on newly collected data, enhancing the robustness and adaptability of the system and ensuring stable operation in various complex environments.

[0104] 5) Reduce costs

[0105] It avoids the high deployment cost of traditional sensor positioning systems and reduces the demand for human resources. By leveraging the existing video surveillance infrastructure, efficient location monitoring can be achieved by simply adding intelligent analysis software modules without the need to install additional expensive hardware equipment, providing a cost-effective solution suitable for large-scale promotion and application.

[0106] 6) Flexibility and scalability

[0107] A modular architecture is designed to support seamless connection with other fire management systems and has good compatibility and scalability. The system interface is standardized and easy to integrate into the existing management system. The configuration can be flexibly adjusted according to needs, which facilitates subsequent maintenance and upgrades. It also provides technical support for possible application scenarios in the future.

[0108] In summary, the present invention has achieved significant improvements in accuracy, degree of automation, response speed, environmental adaptability, cost control and system flexibility through a series of innovative technical means, effectively solving many problems existing in the prior art and providing strong support for safety management in public places.

[0109] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. A method for monitoring the location of fire extinguishers in a hotel environment based on surveillance video, characterized in that: The method is applied to a high-definition camera network, and comprises: S1. Comprehensively collect video surveillance captured by various high-definition cameras in the hotel, obtain real-time video streams and slice the real-time video streams to obtain video slices; S2. Construct a convolutional neural network model, train the convolutional neural network model using the fire extinguisher feature data, and generate a trained feature extraction model; S3, extracting features from the video slices and inputting them into the feature extraction model, comparing them with the fire extinguisher feature data, and generating a comparison result; S4. Use the comparison result to determine whether the current position information of the fire extinguisher conforms to the preset position information. If so, record the current position information and continue the detection. If not, send an alarm and track the movement trajectory of the fire extinguisher.

2. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 1, characterized in that: Comprehensively collect video surveillance captured by various high-definition cameras in the hotel, obtain real-time video streams and slice the real-time video streams to obtain video slices, including: Slice the real-time video stream by frame to obtain video frame slices; The data enhancement algorithm is used to perform data enhancement on the video frame slices to obtain video slices.

3. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 2, characterized in that: Features are extracted from the video slices and input into the feature extraction model, and compared with the fire extinguisher feature data to generate comparison results, including: extracting historical fire extinguisher feature data at the current location from a fire extinguisher feature database; The current fire extinguisher feature data extracted from the video slice is obtained and compared with the historical fire extinguisher feature data at the current location to generate a comparison result.

4. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 3, characterized in that: The fire extinguisher feature data is a multi-scale fire extinguisher feature obtained by fusing fire extinguisher features of different scales.

5. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 4, characterized in that: The multi-scale fire extinguisher features obtained by fusing the fire extinguisher features of different scales include: Acquire original image data and extract low-resolution features, medium-resolution features and high-resolution features from the original image data; Fusing the low-resolution features, medium-resolution features and high-resolution features to form a comprehensive feature vector; The comprehensive feature vector is the multi-scale fire extinguisher feature.

6. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 5, characterized in that: If it does not meet the requirements, an alarm will be sent and the movement of the fire extinguisher will be tracked, including: Compare the current frame and historical data in real time and record the timestamp, and generate the position change data of the fire extinguisher through the comparison results; The position change data and the corresponding timestamp are recorded in the log.

7. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 6, characterized in that: If not met, an alarm is sent and the movement trajectory of the fire extinguisher is tracked, including: Generate an alarm notification by using the fire extinguisher's movement information, location change data and corresponding timestamps; The alarm notification is sent to a security management terminal.

8. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 7, characterized in that: The delivery method includes SMS, email or application push.

9. A method for monitoring the position of a fire extinguisher in a hotel environment based on surveillance video as claimed in claim 8, characterized in that: The method further comprises: The current time and light information in the video stream are obtained, and the current time and light information are combined to perform light compensation on the video frame slices in the video stream.

10. A fire extinguisher location monitoring system in a hotel environment based on surveillance video, characterized in that: The system comprises: The acquisition module is used to comprehensively acquire video surveillance captured by various high-definition cameras in the hotel, obtain real-time video streams and slice the real-time video streams to obtain video slices; A training module, used to construct a convolutional neural network model, train the convolutional neural network model using fire extinguisher feature data, and generate a trained feature extraction model; A comparison module, used for extracting features from the video slices and inputting them into the feature extraction model, comparing them with the fire extinguisher feature data, and generating a comparison result; The detection module is used to use the comparison result to determine whether the current fire extinguisher position information conforms to the preset position information. If it conforms, the current position information is recorded and the detection is continued. If it does not conform, an alarm is sent and the movement trajectory of the fire extinguisher is tracked.

Images