Visual intelligent monitoring alarm system for home-based care for aged

Abstract

The invention discloses a visual intelligent monitoring alarm system for home-based care for the aged, relates to the technical field of intelligent monitoring, and aims to solve the problem that effective safety monitoring and abnormality checking cannot be performed on the living environment of the aged. According to the invention, the early warning intensity can be automatically adjusted, interference of excessive alarm to users is avoided, and important abnormity can be timely and effectively processed. The mechanism for dynamically adjusting the early warning intensity improves the flexibility and the response speed of the system, generates the corresponding alarm type according to the abnormal type of the abnormal monitoring data, and matches the corresponding alarm degree, so that the alarm information is more specific and clear, and through multiple monitoring of face recognition, motion detection, sound analysis and the like, the accuracy of the system is improved. The system can quickly identify abnormal behaviors, trigger an alarm mechanism and ensure the safety of a home-based care environment.

Description

Technical Field

[0001] The present invention relates to the field of smart monitoring technology, and in particular to a visualized smart monitoring and alarm system for home-based elderly care. Background Art

[0002] Smart monitoring refers to a new type of monitoring method that uses modern information technology, especially the Internet of Things, big data, cloud computing, artificial intelligence and other technical means to intelligently manage and control the monitored area.

[0003] The Chinese patent with publication number CN221125337U discloses a fire intelligent monitoring system, which mainly uses a camera terminal and an edge computing terminal, so that when the fire monitoring terminal triggers an alarm, it can combine the real-time situation of the camera terminal as a further judgment, reduce the false triggering action caused by relying solely on the sensor threshold judgment, and thus reduce losses. The traditional image cloud computing method is used to use edge computing terminals for calculation, saving the computing power and storage space of the cloud server, and the response speed is faster. The fire monitoring terminal and the gateway are connected through Bluetooth Mesh network communication. Compared with the traditional electrical connection method, it saves line layout and reduces costs. Compared with traditional wifi and ZigBee, Bluetooth transmission has lower power consumption and a wider transmission distance and faster transmission efficiency, making it suitable for large-scale factory monitoring. Although the above patent solves the problem of fire monitoring, there are still the following problems in actual operation: 1. Failure to use more complete monitoring equipment to effectively monitor the elderly’s monitoring area results in poor security in monitoring the elderly’s living environment.

[0004] 2. The living environment of the elderly is not effectively monitored, and there is no more accurate data comparison of the monitoring data in the monitoring area, resulting in the inability to detect danger in a timely manner.

[0005] 3. No targeted abnormal judgment is made based on the abnormal situation in the monitoring area, resulting in the guardian’s inability to effectively respond to the abnormal situation. Summary of the invention

[0006] The purpose of the present invention is to provide a visual intelligent monitoring alarm system for home-based elderly care, which can automatically adjust the warning intensity, avoiding the interference of excessive alarms on users and ensuring that important anomalies can be handled promptly and effectively. This mechanism of dynamically adjusting the warning intensity improves the flexibility and response speed of the system, generates corresponding alarm types according to the abnormal types of abnormal monitoring data, and matches the corresponding alarm levels, making the alarm information more specific and clear. Through multiple monitoring such as face recognition, motion detection and sound analysis, the system can quickly identify abnormal behavior, trigger the alarm mechanism, ensure the safety of the home-based elderly care environment, and solve the problems in the prior art.

[0007] To achieve the above object, the present invention provides the following technical solutions: Visualized intelligent monitoring and alarm system for home-based elderly care, including: Equipment deployment preparation unit, used to: Deploy monitoring equipment in the monitoring area, perform equipment detection after the deployment of monitoring equipment, and mark the monitoring equipment that has completed the equipment detection as the target monitoring equipment; Monitoring collection and processing unit, used for: The target monitoring equipment is used to acquire monitoring data of the monitoring area, the monitoring data including video data, face recognition data and sound data, and the acquired monitoring data is processed to obtain the target monitoring data; Environmental monitoring unit for: Use target monitoring equipment to monitor environmental data in the monitoring area, and mark the monitored environmental data as target environmental data; Abnormal behavior analysis unit, used to: Perform data anomaly analysis on the target monitoring data and the target environment data respectively, and uniformly mark the anomaly analysis results in the target monitoring data and the target environment data as abnormal monitoring data; Alarm monitoring unit for: The abnormal monitoring data is judged for abnormal type and abnormal degree, alarm data is generated according to the judgment result, and the generated alarm data is sent to the user terminal for alarm information display.

[0008] Preferably, the equipment deployment preparation unit is further used for: The surveillance equipment deployed in the surveillance area includes cameras, environmental sensors, microphones, and facial recognition equipment; Environmental sensors include temperature sensors, humidity sensors, smoke sensors, human motion sensors, and gas sensors; Before the monitoring equipment monitors the monitoring area, each monitoring equipment shall be tested first; Equipment detection includes initialization detection, function detection and alarm trigger detection; After the monitoring equipment has completed the equipment inspection and passed the inspection, it will be marked as the target monitoring equipment; If the monitoring equipment fails the equipment inspection, the monitoring equipment will be repaired or replaced.

[0009] Preferably, the function detection in the equipment detection in the equipment deployment preparation unit includes: An aging determination module, configured to calculate the aging degree of the monitoring device according to the following formula based on the historical usage data of the monitoring device; ; Wherein, L represents the aging degree of the monitoring equipment; Indicates the maximum frequency of use of the monitoring equipment; Indicates the standard frequency of use of monitoring equipment; Indicates the historical usage time of the monitoring device; Indicates the standard usage time of the monitoring equipment; Indicates the historical maintenance times of the monitoring equipment; A position determination module, used to determine a function detection target according to the device type of the monitoring device, and obtain a standard detection range of the function detection target; A judgment module, used to judge whether the actual detection value is within the standard detection range; If so, determining that the functional detection target meets the first requirement; Otherwise, determining that the function detection of the monitoring device of the device type is abnormal; A comprehensive evaluation module, used to calculate the comprehensive evaluation value of the monitoring device of the device type according to the following formula in combination with the aging degree of the monitoring device when all functional detection targets meet the first requirement; ; Wherein, K represents the comprehensive evaluation value of the monitoring equipment of the equipment type; Indicates the number of functional detection targets of the monitoring device; represents a constant, Indicates the aging degree of the monitoring device; Indicates the standard reference test value; represents the actual detection value of the i-th functional detection target of the monitoring device; represents the maximum value of the standard detection range of the i-th functional detection target of the monitoring device; represents the minimum value of the standard detection range of the i-th functional detection target of the monitoring device, Represents the weight of the i-th function detection target; The function detection module is used to determine that the function detection of the monitoring device of the device type is normal when the comprehensive evaluation value of the monitoring device of the device type is greater than a preset evaluation value, otherwise, determine that the function detection of the monitoring device of the device type is abnormal.

[0010] Preferably, the monitoring collection and processing unit includes: Image processing module for: Use the target monitoring equipment to capture video data of the monitoring area; Video data capture is to use a camera to take real-time video images of the monitored area; When the video screen is shot in real time, the video stream transmission function of the camera is started to transmit the real-time shot video screen to the video processing center through the network; The video processing center adds a timestamp to the received video images, decomposes the image frames after the timestamp is added, and performs image enhancement processing on the decomposed image frames; The timestamp of the image frame after image enhancement is matched with the timestamp of the video picture, and after the matching is completed, the video data after the video image enhancement is obtained; Face recognition module for: When a user uses a face recognition device for the first time, the user first collects facial feature data in the face recognition device and is marked as an authorized person after the facial feature data collection is completed; And the face recognition device is connected to the video processing center signal of the camera; The face recognition device decodes the video from the video data from the video processing center and uses the face detection algorithm to detect faces from the video frames; The face detection algorithm recognizes facial features of the faces detected in the video frames; and compare the identified facial features with the facial features of the authorized person; When the feature comparison result is not within the preset threshold range, the face detected in the video data is an unauthorized person; When the feature comparison result is within the preset threshold range, the face detected in the video data is an authorized person; The feature comparison results are uniformly labeled as face recognition data.

[0011] Preferably, the face recognition module compares the recognized facial features with the facial features of the authorized person, including: A feature acquisition module, used for acquiring eye features, mouth features, nose features, eyebrow features, face features, and contour features from the recognized facial features; The feature analysis module is used to determine the distance feature between any two facial features and the overall feature between the facial features based on the eye features, mouth features, nose features, eyebrow features, face features, and contour features obtained in the feature acquisition module; The feature comparison module is used to compare the eye features, mouth features, nose features, eyebrow features and face features with the corresponding part features of the authorized personnel, obtain a single comparison value, and determine whether all the single comparison values ​​are greater than the preset comparison value. If so, determine that the single comparison result passes, otherwise, determine that the single comparison result fails; The feature comparison module is also used to compare the distance feature between any two facial features with the distance feature corresponding to the authorized person when a single comparison result passes, to determine whether the distance difference is within a preset difference range, if so, determine that the distance comparison result passes, otherwise, determine that the distance comparison result fails; The feature comparison module is also used to compare the overall features between the facial features with the overall features corresponding to the authorized person after the distance comparison result is passed, to determine whether the overall feature difference is within a preset overall difference range, if so, determine that the overall comparison result is passed, otherwise, determine that the overall comparison result is not passed; The result recognition module is used to further judge the feature comparison result value based on the comparison value after the single comparison result, the distance comparison result and the overall comparison result are all passed. If any of the single comparison result, the distance comparison result and the overall comparison result fails, it is determined that the face detected in the video data is an unauthorized person.

[0012] Preferably, the result identification module includes: The facial features comparison module is used to calculate the facial features comparison value based on the comparison values ​​of eye features, mouth features, nose features, eyebrow features and facial features, as well as the current recognition and collection status according to the following formula ; ; in, Represents the type of a single identifying feature, =5; j=1, 2, 3, 4, 5 correspond to eye features, mouth features, nose features, eyebrow features and face features respectively; Indicates The difficulty of identifying a single identification feature under the current identification collection situation; Indicates The regional weight of each individual identifying feature; Indicates The comparison value of a single identification feature; Indicates Standard threshold for a single identifying feature; The comprehensive comparison module is used to calculate the feature comparison result value based on the facial features comparison value, combined with the distance comparison result and the overall comparison result according to the following formula ; ; in, Indicates the weight of the five sense organs; represents the distance weight; represents the number of comparisons between any two facial features; s=10; Indicates Distance difference values ​​of facial feature groups; represents the overall weight; represents the overall feature difference value; The personnel determination module is used to determine that when the feature comparison result value is within a preset threshold range, the face detected in the video data is an authorized person, otherwise the face detected in the video data is an unauthorized person.

[0013] Preferably, the monitoring collection and processing unit further includes: Voice recognition module for: Use microphones to collect sound data in the monitored area, and transmit the collected sound data to the sound processing center through the network; The sound processing center optimizes the quality of the received sound data using a sound enhancement algorithm; Perform sound recognition on the sound data after quality optimization; Sound recognition is the identification of voice, sound frequency and sound pattern in sound data; After the sound recognition is completed, a timestamp is performed, and the sound data is obtained after the timestamp is marked; Finally, the video data, face recognition data and sound data are uniformly labeled as target monitoring data.

[0014] Preferably, the environmental monitoring unit is also used for: Use temperature sensors, humidity sensors, smoke sensors, human motion sensors and gas sensors to monitor environmental data in the monitoring area; The temperature sensor is used to monitor the temperature data of the monitoring area; the humidity sensor is used to monitor the humidity data of the monitoring area; the smoke sensor is used to monitor the smoke data of the monitoring area; the human motion sensor is used to monitor whether there are people moving in the monitoring area; the gas sensor is used to monitor the gas concentration in the monitoring area; After the environmental sensor completes the data collection of the monitored area, it will calibrate and verify the data; And mark the monitoring data after calibration and verification with timestamp; After the timestamp is completed, the target environment data is obtained.

[0015] Preferably, the abnormal behavior analysis unit is further used for: First, read the target monitoring data and target environment data respectively; After reading, analyze the target monitoring data for abnormal behavior; Abnormal behavior analysis of target monitoring data includes video data analysis, face recognition analysis and sound data analysis; Video data analysis includes motion detection, behavior recognition, and object detection; Motion detection uses background subtraction to detect abnormal motion in the video, which includes abnormal movement and long-term stillness. Behavior recognition uses convolutional neural networks to detect abnormal behaviors in the video, which include falls and long-term stillness. Object detection uses convolutional neural networks to detect abnormal objects in the video. Face recognition analysis is to analyze unauthorized persons detected in face recognition data and abnormal behaviors of unauthorized persons; Sound data analysis is to convert the speech in the sound data into speech-to-text, and identify abnormal noise and emergency calls based on sound frequency and sound pattern; Then, the target environment data is subjected to an abnormal analysis, and the abnormal analysis of the target environment data is to compare the threshold of the monitoring data of each sensor with the threshold of the standard sensor; The target environment data that is not within the qualified range according to the threshold comparison is marked as abnormal target environment data; Abnormal data analyzed in video data analysis, face recognition analysis and sound data analysis, as well as abnormal target environment data are uniformly marked as abnormal monitoring data.

[0016] Preferably, the alarm monitoring unit is further used for: Read the specific type of the abnormal monitoring data, where the abnormal type is the specific abnormal category of the abnormal monitoring data; Then, the abnormal data analyzed in the video data analysis, face recognition analysis and sound data analysis, as well as the abnormal degree of the abnormal target environment data are distinguished; The degree of abnormality is divided into mild abnormality, moderate abnormality and severe abnormality; Different levels of warning intensity are divided according to the degree of abnormality; Generate corresponding alarm types according to the abnormal types of abnormal monitoring data, including unauthorized personnel, long-term inactivity, fall detection, abnormal noise and voice calls for help; Generate corresponding alarm levels according to alarm types and generate alarm plans based on alarm types and alarm levels; The generated alarm scheme is sent to the user terminal, and the user views the alarm data of the monitored area on the user terminal.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. The visual intelligent monitoring and alarm system for home-based elderly care provided by the present invention can significantly improve the quality of video images through image frame decomposition and image enhancement processing, making the picture clearer and more detailed. The face recognition device can obtain video data from the video processing center in real time, and perform decoding and face detection. The sound processing center uses a sound proliferation algorithm to optimize the quality of the received sound data, which can improve the clarity and recognizability of the sound data. Through multiple monitoring such as face recognition, motion detection and sound analysis, the system can quickly identify abnormal behavior, trigger the alarm mechanism, and ensure the safety of the home-based elderly care environment.

[0018] 2. The visual intelligent monitoring and alarm system for home-based elderly care provided by the present invention can detect abnormalities in environmental data in a timely manner by comparing sensor data with standard thresholds, thereby improving the early warning capability of the system. All analyzed abnormal data are uniformly labeled to facilitate subsequent processing and tracking. The automated and intelligent analysis method greatly improves the processing efficiency of the monitoring system and reduces the need for manual intervention. The efficient collaboration of the sensor network and the data processing system greatly speeds up the collection and processing of environmental data, which is conducive to timely response. By continuously monitoring and analyzing environmental data, abnormal situations can be discovered in a timely manner.

[0019] 3. The visual intelligent monitoring and alarm system for home-based elderly care provided by the present invention can automatically adjust the warning intensity, which not only avoids the interference of excessive alarms on users, but also ensures that important anomalies can be handled promptly and effectively. This mechanism of dynamically adjusting the warning intensity improves the flexibility and response speed of the system, generates corresponding alarm types according to the abnormal types of abnormal monitoring data, and matches the corresponding alarm levels, making the alarm information more specific and clear. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a schematic diagram of the intelligent monitoring alarm unit of the present invention; Figure 2 It is a schematic diagram of the intelligent monitoring alarm process of the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0022] In order to solve the problem that the existing technology does not use more complete monitoring equipment to effectively monitor the monitoring area of ​​the elderly, resulting in poor security of the elderly's living environment monitoring, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions: Visualized intelligent monitoring and alarm system for home-based elderly care, including: Equipment deployment preparation unit, used to: Deploy monitoring equipment in the monitoring area, perform equipment detection after the deployment of monitoring equipment, and mark the monitoring equipment that has completed the equipment detection as the target monitoring equipment; Monitoring collection and processing unit, used for: The target monitoring equipment is used to acquire monitoring data of the monitoring area, the monitoring data including video data, face recognition data and sound data, and the acquired monitoring data is processed to obtain the target monitoring data; Environmental monitoring unit for: Use target monitoring equipment to monitor environmental data in the monitoring area, and mark the monitored environmental data as target environmental data; Abnormal behavior analysis unit, used to: Perform data anomaly analysis on the target monitoring data and the target environment data respectively, and uniformly mark the anomaly analysis results in the target monitoring data and the target environment data as abnormal monitoring data; Alarm monitoring unit for: The abnormal monitoring data is judged for abnormal type and abnormal degree, alarm data is generated according to the judgment result, and the generated alarm data is sent to the user terminal for alarm information display.

[0023] Specifically, by standardizing the equipment inspection process of the equipment deployment preparation unit, once the equipment inspection is qualified, it will be marked as a target monitoring device and can be quickly put into monitoring tasks, thereby improving the response speed and efficiency of the monitoring system. Through multiple monitoring such as face recognition, motion detection and sound analysis of the monitoring acquisition and processing unit, the system can quickly identify abnormal behavior and trigger the alarm mechanism to ensure the safety of the home-based elderly care environment. Through the continuous monitoring and analysis of environmental data by the environmental monitoring unit, abnormal situations can be discovered in time so that preventive measures can be taken. Through the abnormal behavior analysis unit, the sensor data can be compared with the standard threshold, and the abnormalities of the environmental data can be discovered in time, thereby improving the early warning capability of the system. Through the alarm monitoring unit, potential safety hazards can be discovered and dealt with in time, thereby improving the safety performance of the monitoring area.

[0024] The equipment deployment preparation unit is also used to: The surveillance equipment deployed in the surveillance area includes cameras, environmental sensors, microphones, and facial recognition equipment; Environmental sensors include temperature sensors, humidity sensors, smoke sensors, human motion sensors, and gas sensors; Before the monitoring equipment monitors the monitoring area, each monitoring equipment shall be tested first; Equipment detection includes initialization detection, function detection and alarm trigger detection; After the monitoring equipment has completed the equipment inspection and passed the inspection, it will be marked as the target monitoring equipment; If the monitoring equipment fails the equipment inspection, the monitoring equipment will be repaired or replaced.

[0025] Specifically, there are various types of monitoring equipment, including cameras, environmental sensors (temperature, humidity, smoke, human motion and gas sensors), microphones and face recognition equipment, which can comprehensively cover various environments and personnel activities in the monitoring area. Equipment detection is carried out before the monitoring equipment is officially put into use, including initialization detection, function detection and alarm trigger detection, to ensure that the equipment can work normally and prevent monitoring blind spots or false alarms and missed alarms caused by equipment failure. The equipment detection process is standardized. Once the equipment is qualified, it is marked as a target monitoring device and can be quickly put into monitoring tasks to improve the response speed and efficiency of the monitoring system. Qualified monitoring equipment is screened out through equipment detection to ensure the stability and reliability of the monitoring system and reduce the risk of monitoring failure due to equipment problems. Through equipment detection and labeling, the status of each monitoring device can be clearly understood, which is convenient for unified management and scheduling of the monitoring system, and improves the utilization efficiency of monitoring resources. Through timely maintenance and replacement of unqualified equipment, greater losses caused by long-term equipment failures are avoided, which also helps to extend the overall service life of the monitoring equipment and improve cost-effectiveness.

[0026] The functional detection in the device detection in the device deployment preparation unit includes: An aging determination module, configured to calculate the aging degree of the monitoring device according to the following formula based on the historical usage data of the monitoring device; ; Wherein, L represents the aging degree of the monitoring equipment; Indicates the maximum frequency of use of the monitoring equipment; Indicates the standard frequency of use of monitoring equipment; Indicates the historical usage time of the monitoring device; Indicates the standard usage time of the monitoring equipment; Indicates the historical maintenance times of the monitoring equipment; A position determination module, used to determine a function detection target according to the device type of the monitoring device, and obtain a standard detection range of the function detection target; A judgment module, used to judge whether the actual detection value is within the standard detection range; If so, determining that the functional detection target meets the first requirement; Otherwise, determining that the function detection of the monitoring device of the device type is abnormal; A comprehensive evaluation module, used to calculate the comprehensive evaluation value of the monitoring device of the device type according to the following formula in combination with the aging degree of the monitoring device when all functional detection targets meet the first requirement; ; Wherein, K represents the comprehensive evaluation value of the monitoring equipment of the equipment type; Indicates the number of functional detection targets of the monitoring device; represents a constant; Indicates the aging degree of the monitoring device; Indicates the standard reference test value; represents the actual detection value of the i-th functional detection target of the monitoring device; represents the maximum value of the standard detection range of the i-th functional detection target of the monitoring device; represents the minimum value of the standard detection range of the i-th functional detection target of the monitoring device, Represents the weight of the i-th function detection target; The function detection module is used to determine that the function detection of the monitoring device of the device type is normal when the comprehensive evaluation value of the monitoring device of the device type is greater than a preset evaluation value, otherwise, determine that the function detection of the monitoring device of the device type is abnormal.

[0027] In this embodiment, the actual detection value of each functional detection target and its corresponding standard detection range are uniformly standardized and take the value between (0, 3).

[0028] In this embodiment, the greater the influence of the function detection target on the function detection of the monitoring device, the greater the weight of the corresponding function detection target, which is specifically set according to actual conditions.

[0029] The beneficial effect of the above design scheme is: by calculating the aging degree of the monitoring equipment based on the historical usage data of the monitoring equipment, considering the parameters such as usage time, maintenance times, and usage frequency to ensure the accuracy of the obtained aging degree, providing a basis for the functional evaluation of the monitoring equipment, and then determining the function detection target according to the device type of the monitoring equipment, and obtaining the standard detection range of the function detection target, and calculating the comprehensive evaluation value of the monitoring equipment of the device type in combination with the actual detection value and the aging degree. When the comprehensive evaluation value is greater than the preset evaluation value, it is determined that the function detection of the monitoring equipment of the device type is normal, otherwise, it is determined that the function detection of the monitoring equipment of the device type is abnormal, thereby achieving the accuracy of the function detection of the monitoring equipment and providing high-quality monitoring equipment for visual intelligent monitoring of home-based elderly care.

[0030] The monitoring collection and processing unit includes: Image processing module for: Use the target monitoring equipment to capture video data of the monitoring area; Video data capture is to use a camera to take real-time video images of the monitored area; When the video screen is shot in real time, the video stream transmission function of the camera is started to transmit the real-time shot video screen to the video processing center through the network; The video processing center adds a timestamp to the received video images, decomposes the image frames after the timestamp is added, and performs image enhancement processing on the decomposed image frames; The timestamp of the enhanced image frame is matched with the timestamp of the video frame, and after the matching is completed, the video data after the video image enhancement is obtained.

[0031] Specifically, through the video streaming function of the camera, real-time monitoring of the monitoring area is achieved. This helps to detect and respond to abnormal situations in a timely manner, improve safety and emergency response speed. After receiving the video screen, the video processing center will add a timestamp. This step ensures that each video frame has an accurate time record, which is helpful for subsequent video analysis and event tracing. Image frame decomposition and image enhancement processing can significantly improve the quality of video images, making the picture clearer and richer in details. This is crucial for subsequent image recognition, analysis and processing. The divide-and-conquer method of decomposing video data into image frames for processing improves the efficiency of data processing. At the same time, image enhancement processing makes video data more reliable and accurate in subsequent analysis and application. The network transmission and video processing center design make the system easy to integrate with other monitoring equipment and systems. In addition, with the development of technology and the increase in monitoring needs, the system can be easily expanded and upgraded. By optimizing the video data processing process, the solution improves the overall performance of the monitoring system and reduces the false alarm and missed alarm rates caused by poor image quality. This helps save manpower and time costs and improve the economic benefits of the monitoring system.

[0032] Face recognition module for: When a user uses a face recognition device for the first time, the user first collects facial feature data in the face recognition device and is marked as an authorized person after the facial feature data collection is completed; And the face recognition device is connected to the video processing center signal of the camera; The face recognition device decodes the video from the video data from the video processing center and uses the face detection algorithm to detect faces from the video frames; The face detection algorithm performs facial feature recognition on faces detected in video frames; and compare the identified facial features with the facial features of the authorized person; When the feature comparison result is not within the preset threshold range, the face detected in the video data is an unauthorized person; When the feature comparison result is within the preset threshold range, the face detected in the video data is an authorized person; The feature comparison results are uniformly labeled as face recognition data.

[0033] Specifically, by pre-collecting and labeling the facial feature data of authorized personnel, the system can accurately identify authorized and unauthorized personnel. The result of feature comparison determines the access rights, effectively preventing illegal intrusion by unauthorized personnel. The face recognition device can obtain video data from the video processing center in real time, and perform decoding and face detection. The facial feature recognition algorithm can quickly and accurately complete feature comparison, improve the efficiency of identification and verification, and can be easily expanded to more application scenarios. With the continuous advancement of technology, the accuracy and efficiency of the facial feature recognition algorithm will continue to improve, further expanding the possibility of application. By uniformly labeling the feature comparison results as face recognition data, the system can form a complete data record. These data can be used for subsequent analysis, auditing and tracing, improving the standardization and accuracy of data management, combining advanced video processing technology and artificial intelligence algorithms, and realizing intelligent access control management and security monitoring. The system can automatically respond according to preset rules and conditions, such as alarms, logs, etc., to improve the level of intelligent management.

[0034] The face recognition module compares the recognized facial features with the facial features of the authorized person, including: A feature acquisition module, used for acquiring eye features, mouth features, nose features, eyebrow features, face features, and contour features from the recognized facial features; The feature analysis module is used to determine the distance feature between any two facial features and the overall feature between the facial features based on the eye features, mouth features, nose features, eyebrow features, face features, and contour features obtained in the feature acquisition module; The feature comparison module is used to compare the eye features, mouth features, nose features, eyebrow features and face features with the corresponding part features of the authorized personnel, obtain a single comparison value, and determine whether all the single comparison values ​​are greater than the preset comparison value. If so, determine that the single comparison result passes, otherwise, determine that the single comparison result fails; The feature comparison module is also used to compare the distance feature between any two facial features with the distance feature corresponding to the authorized person when a single comparison result passes, to determine whether the distance difference is within a preset difference range, if so, determine that the distance comparison result passes, otherwise, determine that the distance comparison result fails; The feature comparison module is also used to compare the overall features between the facial features with the overall features corresponding to the authorized person after the distance comparison result is passed, to determine whether the overall feature difference is within a preset overall difference range, if so, determine that the overall comparison result is passed, otherwise, determine that the overall comparison result is not passed; The result recognition module is used to further judge the feature comparison result value based on the comparison value after the single comparison result, the distance comparison result and the overall comparison result are all passed. If any of the single comparison result, the distance comparison result and the overall comparison result fails, it is determined that the face detected in the video data is an unauthorized person.

[0035] In this embodiment, the feature comparison result value is further judged as follows: when the feature comparison result value is within a preset threshold range, the face detected in the video data is an authorized person, otherwise the face detected in the video data is an unauthorized person.

[0036] The beneficial effect of the above design scheme is: by comparing the facial features of the face step by step from three aspects: facial features, the distance between the facial features and the overall appearance, and making another comprehensive judgment after the comparison, the accuracy and efficiency of facial feature recognition can be achieved, the illegal intrusion of unauthorized persons can be effectively prevented, the security of visual intelligent monitoring of home-based elderly care can be ensured, and customer satisfaction can be improved.

[0037] The result identification module comprises: The facial features comparison module is used to calculate the facial features comparison value based on the comparison values ​​of eye features, mouth features, nose features, eyebrow features and facial features, as well as the current recognition and collection status according to the following formula ; ; in, Represents the type of a single identifying feature; =5; j=1, 2, 3, 4, 5 correspond to eye features, mouth features, nose features, eyebrow features and face features respectively; Indicates The difficulty of identifying a single identification feature under the current identification collection situation; Indicates The regional weight of each individual identifying feature; Indicates The comparison value of a single identification feature; Indicates Standard threshold for a single identifying feature; The comprehensive comparison module is used to calculate the feature comparison result value based on the facial features comparison value, combined with the distance comparison result and the overall comparison result according to the following formula ; ; in, Indicates the weight of the five sense organs; represents the distance weight; represents the number of comparisons between any two facial features; s=10; Indicates Distance difference values ​​of facial feature groups; represents the overall weight; represents the overall feature difference value; The personnel determination module is used to determine that when the feature comparison result value is within a preset threshold range, the face detected in the video data is an authorized person, otherwise the face detected in the video data is an unauthorized person.

[0038] In this embodiment, the comparison values, difference values ​​and weights of each type are standardized in advance, eliminating the differences caused by different types, value units, etc.

[0039] In this embodiment, the recognition difficulty in the current recognition acquisition situation is related to the face angle of the acquired image, etc.

[0040] In this embodiment, the design of the weights is determined in advance according to actual conditions.

[0041] The beneficial effects of the above design scheme are: by comprehensively determining the quantitative comparison results of the facial features based on the current recognition and collection situation and the facial features comparison results, and then determining the final feature comparison result value based on the weight setting, the quantification of recognition is achieved, the efficiency and accuracy of face recognition are ensured, and the illegal intrusion of unauthorized personnel is effectively prevented. The security of visual intelligent monitoring of home-based elderly care is ensured, and customer satisfaction is improved.

[0042] Voice recognition module for: Use microphones to collect sound data in the monitored area, and transmit the collected sound data to the sound processing center through the network; The sound processing center optimizes the quality of the received sound data using a sound enhancement algorithm; Perform sound recognition on the sound data after quality optimization; Sound recognition is the identification of voice, sound frequency and sound pattern in sound data; After the sound recognition is completed, a timestamp is performed, and the sound data is obtained after the timestamp is marked; Finally, the video data, face recognition data and sound data are uniformly labeled as target monitoring data.

[0043] Specifically, using microphones to collect sound data can capture the sound information of the monitored area in real time, ensuring the timeliness and integrity of the data. As a common sound collection device, microphones are easy to obtain and relatively low in cost, which is convenient for large-scale deployment and application. The collected sound data is transmitted to the sound processing center through the network, realizing remote transmission and centralized processing of data, and improving processing efficiency. Network transmission is characterized by high speed and stability, which can ensure the real-time and accuracy of sound data. The sound processing center uses the sound proliferation algorithm to optimize the quality of the received sound data, which can improve the clarity and recognizability of the sound data. The application of the sound proliferation algorithm helps to reduce noise interference and improve the overall quality of sound data. Sound recognition technology can accurately identify the voice, sound frequency and sound pattern in the sound data, providing a reliable foundation for subsequent data analysis and application. The continuous improvement of sound recognition technology makes the recognition results more accurate and reliable. Timestamping the sound data can accurately record the time point when the sound event occurs, which is convenient for subsequent data backtracking and analysis. The accuracy of timestamp annotation helps to improve the efficiency and accuracy of data processing. Through multiple monitoring such as face recognition, motion detection and sound analysis, the system can quickly identify abnormal behavior, trigger the alarm mechanism, and ensure the safety of the home elderly care environment.

[0044] In order to solve the problem that the existing technology does not effectively monitor the living environment of the elderly and does not compare the monitoring data of the monitoring area more accurately, which leads to the failure to detect danger in time, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions: Environmental monitoring unit, also used for: Use temperature sensors, humidity sensors, smoke sensors, human motion sensors and gas sensors to monitor environmental data in the monitoring area; The temperature sensor is used to monitor the temperature data of the monitoring area; the humidity sensor is used to monitor the humidity data of the monitoring area; the smoke sensor is used to monitor the smoke data of the monitoring area; the human motion sensor is used to monitor whether there are people moving in the monitoring area; the gas sensor is used to monitor the gas concentration in the monitoring area; After the environmental sensor completes the data collection of the monitored area, it will calibrate and verify the data; And mark the monitoring data after calibration and verification with timestamp; After the timestamp is completed, the target environment data is obtained.

[0045] Specifically, through a variety of sensors such as temperature, humidity, smoke, human motion and gas, the environmental parameters of the monitored area can be fully monitored to ensure the diversity and integrity of the data. The sensors can collect data in real time and perform calibration and verification immediately so that the environmental status of the monitored area can be reflected in a timely manner. The data calibration and verification steps ensure the accuracy of the monitoring data and reduce erroneous data caused by sensor errors or external interference. Timestamp annotation ensures that each piece of monitoring data has a clear time record, which is convenient for subsequent data analysis and traceability. In places where security needs to be monitored (such as warehouses, data centers, etc.), human motion sensors can monitor intruders in real time, while smoke and gas sensors can provide early warnings in the event of fire or harmful gas leaks, thereby improving safety. Steps such as data collection, calibration, verification and timestamp annotation can usually be performed by automated systems, reducing manual intervention and improving ease of use. The efficient collaboration of sensor networks and data processing systems greatly speeds up the collection and processing of environmental data, which is conducive to timely response. By continuously monitoring and analyzing environmental data, abnormal situations can be discovered in a timely manner, so that preventive measures can be taken to avoid potential environmental risks.

[0046] The abnormal behavior analysis unit is also used to: First, read the target monitoring data and target environment data respectively; After reading, analyze the target monitoring data for abnormal behavior; Abnormal behavior analysis of target monitoring data includes video data analysis, face recognition analysis and sound data analysis; Video data analysis includes motion detection, behavior recognition, and object detection; Motion detection uses background subtraction to detect abnormal motion in the video, which includes abnormal movement and long-term stillness. Behavior recognition uses convolutional neural networks to detect abnormal behaviors in the video, which include falls and long-term stillness. Object detection uses convolutional neural networks to detect abnormal objects in the video. Face recognition analysis is to analyze unauthorized persons detected in face recognition data and abnormal behaviors of unauthorized persons; Sound data analysis is to convert the speech in the sound data into speech-to-text, and identify abnormal noise and emergency calls based on sound frequency and sound pattern; Then, the target environment data is subjected to an abnormal analysis, and the abnormal analysis of the target environment data is to compare the threshold of the monitoring data of each sensor with the threshold of the standard sensor; The target environment data that is not within the qualified range according to the threshold comparison is marked as abnormal target environment data; Abnormal data analyzed in video data analysis, face recognition analysis and sound data analysis, as well as abnormal target environment data are uniformly marked as abnormal monitoring data.

[0047] Specifically, for video data, the solution not only performs motion detection, behavior recognition and object detection, but also considers a variety of abnormal behaviors such as abnormal movement, long-term stillness, and falls, which improves the accuracy and meticulousness of the analysis. Face recognition analysis can identify unauthorized personnel and their abnormal behaviors, enhancing security. Sound data analysis further improves the responsiveness of the monitoring system through voice-to-text conversion and abnormal noise and emergency call recognition. Advanced artificial intelligence technologies such as convolutional neural networks are used for behavior recognition and object detection, improving the automation and intelligence level of analysis. The application of technologies such as background subtraction makes motion detection more accurate and efficient. By comparing sensor data with standard thresholds, environmental data anomalies can be discovered in a timely manner, improving the system's early warning capability. All analyzed abnormal data are uniformly labeled for subsequent processing and tracking. The automated and intelligent analysis method greatly improves the processing efficiency of the monitoring system and reduces the need for manual intervention. The rapid labeling and identification of abnormal data helps to quickly respond to and handle potential security issues. By collecting and analyzing a large amount of monitoring data, it can provide strong data support for decision-making and improve the accuracy and scientificity of decision-making.

[0048] In order to solve the problem in the prior art that there is no targeted abnormal judgment based on the abnormal situation in the monitoring area, which results in the inability of the guardian to effectively respond to the abnormal situation, please refer to Figure 1 and Figure 2 , this embodiment provides the following technical solutions: The alarm monitoring unit is also used to: Read the specific type of the abnormal monitoring data, where the abnormal type is the specific abnormal category of the abnormal monitoring data; Then, the abnormal data analyzed in the video data analysis, face recognition analysis and sound data analysis, as well as the abnormal degree of the abnormal target environment data are distinguished; The degree of abnormality is divided into mild abnormality, moderate abnormality and severe abnormality; Different levels of warning intensity are divided according to the degree of abnormality; Generate corresponding alarm types according to the abnormal types of abnormal monitoring data, including unauthorized personnel, long-term inactivity, fall detection, abnormal noise and voice calls for help; Generate corresponding alarm levels according to alarm types and generate alarm plans based on alarm types and alarm levels; The generated alarm scheme is sent to the user terminal, and the user views the alarm data of the monitored area on the user terminal.

[0049] Specifically, by subdividing abnormal monitoring data into specific abnormal categories, the system can more accurately identify and handle different types of abnormal situations, thereby improving the precision and accuracy of monitoring. By performing abnormal analysis on different types of monitoring data, such as video data, face recognition data, and sound data, the monitoring area can be more comprehensively covered, reducing the possibility of missed reports and false reports. According to the degree of abnormality (minor abnormality, moderate abnormality, and severe abnormality), the system can automatically adjust the warning intensity, which not only avoids excessive alarms from interfering with users, but also ensures that important abnormalities can be handled in a timely and effective manner. This mechanism of dynamically adjusting the warning intensity improves the flexibility and response speed of the system. According to the abnormal type of abnormal monitoring data, the corresponding alarm type (such as unauthorized entry, long-term inactivity, fall detection, abnormal noise, and voice calls for help) is generated, and the corresponding alarm degree is matched, making the alarm information more specific and clear. This precise matching helps users quickly understand the alarm content and take corresponding countermeasures. The generated alarm plan is sent to the user terminal, and users can easily view the alarm data of the monitoring area on the user terminal, thereby improving the ease of use and user satisfaction of the system. Users can respond promptly based on the alarm information, effectively reducing potential safety risks. New abnormal categories, alarm types, and alarm levels can be added according to actual needs to meet monitoring needs in different scenarios. At the same time, the system is also highly flexible and can optimize and upgrade functions based on user feedback and actual needs. Through real-time monitoring and abnormality detection, the system can promptly discover and deal with potential safety hazards, improving the safety performance of the monitored area. For some emergency abnormal situations (such as voice calls for help), the system can quickly issue an alarm to buy valuable time for rescue work.

[0050] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0051] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

Claims

1. Visualized intelligent monitoring and alarm system for home-based elderly care, characterized by: include: Equipment deployment preparation unit, used to: Deploy monitoring equipment in the monitoring area, perform equipment detection after the deployment of monitoring equipment, and mark the monitoring equipment that has completed the equipment detection as the target monitoring equipment; Monitoring collection and processing unit, used for: The target monitoring equipment is used to acquire monitoring data of the monitoring area, the monitoring data including video data, face recognition data and sound data, and the acquired monitoring data is processed to obtain the target monitoring data; Environmental monitoring unit for: Use target monitoring equipment to monitor environmental data in the monitoring area, and mark the monitored environmental data as target environmental data; Abnormal behavior analysis unit, used to: Perform data anomaly analysis on the target monitoring data and the target environment data respectively, and uniformly mark the anomaly analysis results in the target monitoring data and the target environment data as abnormal monitoring data; Alarm monitoring unit for: The abnormal monitoring data is judged for abnormal type and abnormal degree, alarm data is generated according to the judgment result, and the generated alarm data is sent to the user terminal for alarm information display.

2. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 1 is characterized in that: The equipment deployment preparation unit is further used for: The surveillance equipment deployed in the surveillance area includes cameras, environmental sensors, microphones, and facial recognition equipment; Environmental sensors include temperature sensors, humidity sensors, smoke sensors, human motion sensors, and gas sensors; Before the monitoring equipment monitors the monitoring area, each monitoring equipment shall be tested first; Equipment detection includes initialization detection, function detection and alarm trigger detection; After the monitoring equipment has completed the equipment inspection and passed the inspection, it will be marked as the target monitoring equipment; If the monitoring equipment fails the equipment inspection, the monitoring equipment will be repaired or replaced.

3. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 2 is characterized in that: The functional detection in the device detection in the device deployment preparation unit includes: An aging determination module, configured to calculate the aging degree of the monitoring device according to the following formula based on the historical usage data of the monitoring device; ； Wherein, L represents the aging degree of the monitoring equipment; Indicates the maximum frequency of use of the monitoring equipment; Indicates the standard frequency of use of monitoring equipment; Indicates the historical usage time of the monitoring device; Indicates the standard usage time of the monitoring equipment; Indicates the historical maintenance times of the monitoring equipment; A position determination module, used to determine a function detection target according to the device type of the monitoring device, and obtain a standard detection range of the function detection target; A judgment module, used to judge whether the actual detection value is within the standard detection range; If yes, determining that the functional detection target meets the first requirement; Otherwise, determining that the function detection of the monitoring device of the device type is abnormal; A comprehensive evaluation module, used to calculate the comprehensive evaluation value of the monitoring device of the device type according to the following formula in combination with the aging degree of the monitoring device when all functional detection targets meet the first requirement; ； Wherein, K represents the comprehensive evaluation value of the monitoring equipment of the equipment type; Indicates the number of functional detection targets of the monitoring device; represents a constant; Indicates the aging degree of the monitoring device; Indicates the standard reference test value; represents the actual detection value of the i-th functional detection target of the monitoring device; represents the maximum value of the standard detection range of the i-th functional detection target of the monitoring device; represents the minimum value of the standard detection range of the i-th functional detection target of the monitoring device, Represents the weight of the i-th function detection target; The function detection module is used to determine that the function detection of the monitoring device of the device type is normal when the comprehensive evaluation value of the monitoring device of the device type is greater than a preset evaluation value, otherwise, determine that the function detection of the monitoring device of the device type is abnormal.

4. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 2 is characterized in that: The monitoring collection and processing unit comprises: Image processing module for: Use the target monitoring equipment to capture video data of the monitoring area; Video data capture is to use a camera to take real-time video images of the monitored area; When the video screen is shot in real time, the video stream transmission function of the camera is started to transmit the real-time shot video screen to the video processing center through the network; The video processing center adds a timestamp to the received video images, decomposes the image frames after the timestamp is added, and performs image enhancement processing on the decomposed image frames; The timestamp of the image frame after image enhancement is matched with the timestamp of the video picture, and after the matching is completed, the video data after the video image enhancement is obtained; Face recognition module for: When a user uses a face recognition device for the first time, the user first collects facial feature data in the face recognition device and is marked as an authorized person after the facial feature data collection is completed; And the face recognition device is connected to the video processing center signal of the camera; The face recognition device decodes the video from the video data from the video processing center and uses the face detection algorithm to detect faces from the video frames; The face detection algorithm recognizes facial features of the faces detected in the video frames; and compare the identified facial features with the facial features of the authorized person; When the feature comparison result is not within the preset threshold range, the face detected in the video data is an unauthorized person; When the feature comparison result is within the preset threshold range, the face detected in the video data is an authorized person; The feature comparison results are uniformly labeled as face recognition data.

5. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 4 is characterized in that: The face recognition module compares the recognized facial features with the facial features of the authorized person, including: A feature acquisition module, used for acquiring eye features, mouth features, nose features, eyebrow features, face features, and contour features from the recognized facial features; The feature analysis module is used to determine the distance feature between any two facial features and the overall feature between the facial features based on the eye features, mouth features, nose features, eyebrow features, face features, and contour features obtained in the feature acquisition module; The feature comparison module is used to compare the eye features, mouth features, nose features, eyebrow features and face features with the corresponding part features of the authorized personnel, obtain a single comparison value, and determine whether all the single comparison values ​​are greater than the preset comparison value. If so, determine that the single comparison result passes, otherwise, determine that the single comparison result fails; The feature comparison module is also used to compare the distance feature between any two facial features with the distance feature corresponding to the authorized person when a single comparison result passes, to determine whether the distance difference is within a preset difference range, if so, determine that the distance comparison result passes, otherwise, determine that the distance comparison result fails; The feature comparison module is also used to compare the overall features between the facial features with the overall features corresponding to the authorized person after the distance comparison result is passed, to determine whether the overall feature difference is within a preset overall difference range, if so, determine that the overall comparison result is passed, otherwise, determine that the overall comparison result is not passed; The result recognition module is used to further judge the feature comparison result value based on the comparison value after the single comparison result, the distance comparison result and the overall comparison result are all passed. If any of the single comparison result, the distance comparison result and the overall comparison result fails, it is determined that the face detected in the video data is an unauthorized person.

6. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 5 is characterized in that: The result identification module comprises: The facial features comparison module is used to calculate the facial features comparison value based on the comparison values ​​of eye features, mouth features, nose features, eyebrow features and facial features, as well as the current recognition and collection status according to the following formula ; ； in, Represents the type of a single identifying feature, =5; j=1, 2, 3, 4, 5 correspond to eye features, mouth features, nose features, eyebrow features and face features respectively; Indicates The difficulty of identifying a single identification feature under the current identification collection situation; Indicates The regional weight of a single identification feature, Indicates The comparison value of a single identification feature; Indicates Standard threshold for a single identifying feature; The comprehensive comparison module is used to calculate the feature comparison result value based on the facial features comparison value, combined with the distance comparison result and the overall comparison result according to the following formula ; ； in, Indicates the weight of the five sense organs; represents the distance weight; represents the number of comparisons between any two facial features; s=10; Indicates Distance difference values ​​of facial feature groups; represents the overall weight; represents the overall feature difference value; The personnel determination module is used to determine that when the feature comparison result value is within a preset threshold range, the face detected in the video data is an authorized person, otherwise the face detected in the video data is an unauthorized person.

7. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 6 is characterized in that: The monitoring collection and processing unit further includes: Voice recognition module for: Use microphones to collect sound data in the monitored area, and transmit the collected sound data to the sound processing center through the network; The sound processing center optimizes the quality of the received sound data using a sound enhancement algorithm; Perform sound recognition on the sound data after quality optimization; Sound recognition is the identification of voice, sound frequency and sound pattern in sound data; After the sound recognition is completed, a timestamp is performed, and the sound data is obtained after the timestamp is marked; Finally, the video data, face recognition data and sound data are uniformly labeled as target monitoring data.

8. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 7 is characterized in that: The environmental monitoring unit is also used for: Use temperature sensors, humidity sensors, smoke sensors, human motion sensors and gas sensors to monitor environmental data in the monitoring area; The temperature sensor is used to monitor the temperature data of the monitoring area; the humidity sensor is used to monitor the humidity data of the monitoring area; the smoke sensor is used to monitor the smoke data of the monitoring area; the human motion sensor is used to monitor whether there are people moving in the monitoring area; the gas sensor is used to monitor the gas concentration in the monitoring area; After the environmental sensor completes the data collection of the monitored area, it will calibrate and verify the data; And mark the monitoring data after calibration and verification with timestamp; After the timestamp is completed, the target environment data is obtained.

9. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 8 is characterized in that: The abnormal behavior analysis unit is further used for: First, read the target monitoring data and target environment data respectively; After reading, analyze the target monitoring data for abnormal behavior; Abnormal behavior analysis of target monitoring data includes video data analysis, face recognition analysis and sound data analysis; Video data analysis includes motion detection, behavior recognition, and object detection; Motion detection uses background subtraction to detect abnormal motion in the video, which includes abnormal movement and long-term stillness. Behavior recognition uses convolutional neural networks to detect abnormal behaviors in the video, which include falls and long-term stillness. Object detection uses convolutional neural networks to detect abnormal objects in the video. Face recognition analysis is to analyze unauthorized persons detected in face recognition data and abnormal behaviors of unauthorized persons; Sound data analysis is to convert the speech in the sound data into speech-to-text, and identify abnormal noise and emergency calls based on sound frequency and sound pattern; Then, the target environment data is subjected to an abnormal analysis, and the abnormal analysis of the target environment data is to compare the threshold of the monitoring data of each sensor with the threshold of the standard sensor; The target environment data that is not within the qualified range according to the threshold comparison is marked as abnormal target environment data; Abnormal data analyzed in video data analysis, face recognition analysis and sound data analysis, as well as abnormal target environment data are uniformly marked as abnormal monitoring data.

10. The visual intelligent monitoring and alarm system for home-based elderly care according to claim 9 is characterized in that: The alarm monitoring unit is further used for: Read the specific type of the abnormal monitoring data, where the abnormal type is the specific abnormal category of the abnormal monitoring data; Then, the abnormal data analyzed in the video data analysis, face recognition analysis and sound data analysis, as well as the abnormal degree of the abnormal target environment data are distinguished; The degree of abnormality is divided into mild abnormality, moderate abnormality and severe abnormality; Different levels of warning intensity are divided according to the degree of abnormality; Generate corresponding alarm types according to the abnormal types of abnormal monitoring data, including unauthorized personnel, long-term inactivity, fall detection, abnormal noise and voice calls for help; Generate corresponding alarm levels according to alarm types and generate alarm plans based on alarm types and alarm levels; The generated alarm scheme is sent to the user terminal, and the user views the alarm data of the monitored area on the user terminal.

Images