A gas self-closing valve control method and system, an intelligent terminal and a storage medium

Abstract

The application relates to a gas self-closing valve control method and system, an intelligent terminal and a storage medium, relates to the field of gas self-closing valve technology, and the method comprises the following steps: collecting a room image; determining a target image from the room image according to the position of a gas self-closing valve corresponding gas equipment in a house; collecting personnel identity and personnel behavior corresponding to the personnel identity from the target image in the case that the actual opening degree of the gas self-closing valve is greater than a preset opening degree, wherein the personnel identity comprises a self-awareness level; and adjusting the opening degree of the gas self-closing valve according to the personnel identity and the personnel behavior. The application has the effect of improving the safety of the gas self-closing valve.

Description

Technical Field

[0001] This application relates to the field of gas self-closing valve technology, and in particular to a gas self-closing valve control method, system, smart terminal and storage medium. Background Technology

[0002] As a key device for household gas safety protection, the gas self-closing valve can automatically cut off the gas supply when the gas pressure is abnormal, the pipeline leaks, or the gas supply is interrupted, effectively preventing gas accidents caused by equipment failure or pipeline damage.

[0003] In related technologies, the control of gas self-closing valves usually relies on the threshold judgment of a single physical parameter such as a pressure sensor, flow sensor, or gas concentration alarm, and directly triggers the valve closing operation when an abnormality is detected.

[0004] For the aforementioned technologies, relying solely on physical parameter thresholds or the presence or absence of personnel for judgment can easily lead to misjudgments or omissions. Summary of the Invention

[0005] To improve the safety of gas self-closing valves, this application provides a gas self-closing valve control method, system, smart terminal, and storage medium.

[0006] In a first aspect, this application provides a method for controlling a gas self-closing valve, which adopts the following technical solution: A method for controlling a gas self-closing valve includes: Capture room images; The target image is determined from the room image based on the location of the gas appliance corresponding to the gas self-closing valve in the house; When the actual opening of the gas self-closing valve is greater than the preset opening, the personnel identity and the corresponding personnel behavior are collected from the target image, and the personnel identity includes the level of self-awareness. Adjust the opening degree of the gas self-closing valve according to the identity and behavior of the personnel.

[0007] By adopting the above technical solution, by acquiring room images and determining the target image corresponding to the gas equipment, and with the gas self-closing valve in the open state, the identity of the personnel and their corresponding behaviors are collected from the target image. The personnel identity includes their level of self-awareness. Then, the opening degree of the gas self-closing valve is adjusted according to the personnel identity and behavior. This makes the control of the gas self-closing valve no longer solely rely on physical parameter thresholds such as pressure, flow rate, or concentration, but can make comprehensive decisions based on the cognitive characteristics and actual behavior of the people in the kitchen. This realizes the transformation from indiscriminate passive cut-off to proactive differentiated protection based on personnel perception and behavioral understanding, effectively balancing gas safety and user experience.

[0008] Optionally, based on the self-awareness level in the personnel identity, the personnel in the personnel image are divided into a first person and a second person, wherein the self-awareness level of the first person is greater than a preset threshold, and the self-awareness level of the second person is less than or equal to the preset threshold. If the personnel image only includes the first personnel, then the opening degree of the gas self-closing valve is adjusted according to the personnel behavior and the first dangerous behavior; If the personnel image includes at least one first person and at least one second person, then the degree of control of the first person over the behavior of the second person is generated, the degree of control of the behavior is used to describe the degree of intervention in the behavior of other persons; the opening degree of the gas self-closing valve is adjusted according to the degree of control of the behavior and the behavior of the personnel. If the personnel image only includes the second person, then the opening degree of the gas self-closing valve is adjusted according to the personnel behavior and the second dangerous behavior, where the danger level of the second dangerous behavior is greater than the danger level of the first dangerous behavior.

[0009] By adopting the above technical solution, and by collecting the self-awareness level of personnel from the target image, the personnel in the kitchen are divided into first personnel and second personnel. Different opening adjustment strategies are implemented according to different personnel combination scenarios. This realizes the technical upgrade from a one-size-fits-all valve closure to graded protection based on personnel's cognitive ability. It avoids frequent accidental valve closure interference for personnel with self-awareness and strengthens the proactive protection for personnel who lack the ability to identify danger.

[0010] Optionally, based on the target image, the first person's first person features and the second person's second person features are obtained; Based on the first personnel characteristics and the second personnel characteristics, the initial degree of control is obtained; From the personnel behaviors, obtain the first personnel behavior corresponding to the first personnel and the interaction state between the first personnel and the second personnel; Based on the behavior of the first person and the interaction state, a correction coefficient is generated; The behavioral control degree is obtained by multiplying the initial control degree by the correction coefficient.

[0011] By adopting the above technical solution, when a first person and a second person are present in the kitchen at the same time, the initial degree of control is obtained by acquiring the characteristics of both persons, and a correction coefficient is generated by combining the behavior and interaction status of the first person. Finally, the behavioral control degree is calculated in the form of a product, which realizes the dynamic quantitative assessment of the guardian's actual intervention ability. This allows the valve closing decision to truly reflect the degree of protection of the second person in the current scenario, reducing the false alarm rate and the missed alarm rate.

[0012] Optionally, the gaze behavior features of the first person towards the second person can be extracted from the behavior of the first person. The gaze behavior features include the angle between the gaze direction of the first person and the gaze direction of the line connecting the first person and the second person. Extract the degree of attention the first person pays to the second person from the interaction state; Physical distance information between the first person and the second person is extracted from the target image. The physical distance information is used to characterize whether there is an object between them that prevents the first person from interfering with the behavior of the second person. Based on the physical interval information, determine the interval penalty coefficient; The correction coefficient is generated based on the gaze behavior characteristics, the level of attention, and the interval penalty coefficient.

[0013] By adopting the above technical solution, when generating the correction coefficient, the angle of the first person's gaze, the degree of attention, and the physical distance between them are comprehensively extracted, and the physical distance is converted into an interval penalty coefficient and included in the calculation of the correction coefficient. This achieves a multi-dimensional integrated judgment of the guardian's attention allocation, attention intensity, and physical accessibility, effectively avoiding the problem of overestimation of intervention ability due to physical obstacles such as stoves and islands, and improving the accuracy of control assessment.

[0014] Optionally, the second person's subdivided identity type is identified, the subdivided identity type including at least a first subdivided type and a second subdivided type, the first subdivided type corresponding to a child and the second subdivided type corresponding to an elderly person; When the second person belongs to the first sub-category, the relative spatial parameters between the second person and the gas equipment are extracted from the person's behavior. The relative spatial parameters include real-time distance and direction of movement. If the real-time distance and the direction of movement meet the first determination condition, the person's behavior is determined to be the second dangerous behavior. When the second person belongs to the second sub-type, the operation timing features of the second person are extracted from the person's behavior; if the operation timing features match the preset abnormal behavior pattern, the person's behavior is determined to be the second dangerous behavior.

[0015] By adopting the above technical solution, in high-risk scenarios where only a second person is present, the system can further identify the second person's specific identity type, such as whether they are a child or an elderly person. For different specific types, the system uses judgment conditions based on spatial parameters and abnormal behavior pattern matching based on operation sequence characteristics. This achieves targeted adaptation to the behavioral characteristics of different vulnerable groups, enabling the security protection strategy to fit the actual risk performance of each group and improving the accuracy of dangerous behavior identification.

[0016] Optionally, the real-time distance and direction of movement of the second person can be obtained; When the real-time distance has entered a preset high-risk zone, the direction of movement is toward the gas equipment, and the real-time distance continues to decrease within a preset time period, the person's behavior is determined to be the second dangerous behavior. When the real-time distance and the direction of movement are not determined to be a second dangerous behavior, the system detects whether the knob of the gas device is in the on state and whether no effective ignition behavior is detected in the target image. If so, the distance between any part of the second person's body and the knob is obtained. When the distance is continuously less than a preset first contact threshold, the person's behavior is determined to be the second dangerous behavior. When the distance does not determine the second dangerous behavior, the body center height and body center projection position of the second person are extracted from the target image; When the height of the body's center of gravity exceeds a preset safe height threshold, and the projected position of the body's center of gravity falls into a preset restricted area above the gas equipment, the person's behavior is determined to be the second dangerous behavior. When the body's center of gravity height and the body's center of gravity projection position are not determined to be a second dangerous behavior, the system detects whether the flexible pipe connected to the gas equipment has undergone displacement exceeding a preset deformation. If so, determine whether there is a spatial correlation between the time of the displacement and the limb movement trajectory of the second person; When the spatial correlation exists, the person's behavior is determined to be the second dangerous behavior.

[0017] By adopting the above technical solution, for children as the second type of person, a progressive investigation is carried out from four aspects: real-time distance and direction of movement, contact distance between body parts and knobs when the knobs are turned on, height of body center of gravity and projection position, and abnormal displacement of flexible pipes. This covers a variety of common accidental behaviors of children, such as actively approaching and touching knobs, climbing stoves and pulling gas pipes, forming a complete behavioral risk protection mechanism and reducing the possibility of missed detection.

[0018] Optionally, after detecting that the gas equipment has been started, a first preset time window is started for timing; When the second person is detected leaving the monitoring area covered by the target image within the first preset time window, and the time spent leaving reaches the second preset time window, it is determined that the person's behavior matches the abnormal behavior pattern. When the duration of the departure does not match, while the gas equipment remains running, it is detected whether the relative stillness of the second person in the monitoring area lasts for more than a third preset time window. If so, determine whether a predetermined action sequence related to cooking operations has been identified from the target image within the third preset time window; When the predetermined action sequence is not identified, it is determined that the person's behavior matches the abnormal behavior pattern; When the predetermined action sequence is detected, the number of times the second person performs the on-off cycle on the operating components of the gas equipment is counted within a preset statistical period; When the number of on-off cycles exceeds a preset number, and no corresponding valid ignition event is detected in the target image after each on-off operation, the personnel behavior is determined to match the abnormal behavior pattern. When the number of on-off cycles does not match, detect from the target image whether the second person has placed an item on a running gas appliance; If so, determine whether the item belongs to a preset cooking utensil category; When the item does not belong to the cooking utensil category, the person's behavior is determined to match the abnormal behavior pattern.

[0019] By adopting the above technical solution, for the second type of elderly people, the following behaviors were detected in sequence: the duration of time after the gas appliance is started and the person leaves the monitoring area; the duration of time in front of the stove where the person is relatively still and not cooking; the correspondence between the number of on-off cycles and effective ignition events; and the behavior of placing non-cooking items on the running gas appliance. This covered typical cognitive impairment behaviors such as forgetting to start the stove, inability to use the stove, repeated ineffective operations, and placement of abnormal items, thus achieving in-depth identification of gas usage risks for the elderly.

[0020] Secondly, this application provides a gas self-closing valve control system, which adopts the following technical solution: A gas self-closing valve control system includes: The acquisition module is used to acquire room images; and to determine the target image from the room images based on the location of the gas appliance corresponding to the gas self-closing valve in the house. The calculation module is used to collect the identity of the person and the corresponding behavior of the person from the target image when the actual opening degree of the gas self-closing valve is greater than the preset opening degree. The identity of the person includes the level of self-awareness. The processing module is used to adjust the opening degree of the gas self-closing valve according to the identity and behavior of the personnel.

[0021] By adopting the above technical solution, by acquiring room images and determining the target image corresponding to the gas equipment, and with the gas self-closing valve in the open state, the identity of the personnel and their corresponding behaviors are collected from the target image. The personnel identity includes their level of self-awareness. Then, the opening degree of the gas self-closing valve is adjusted according to the personnel identity and behavior. This makes the control of the gas self-closing valve no longer solely rely on physical parameter thresholds such as pressure, flow rate, or concentration, but can make comprehensive decisions based on the cognitive characteristics and actual behavior of the people in the kitchen. This realizes the transformation from indiscriminate passive cut-off to proactive differentiated protection based on personnel perception and behavioral understanding, effectively balancing gas safety and user experience.

[0022] Thirdly, this application provides a smart terminal, which adopts the following technical solution: A smart terminal includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute the method described in any one of the above.

[0023] Fourthly, this application provides a computer storage medium capable of storing corresponding programs, which facilitates the improvement of the safety of gas self-closing valves, and adopts the following technical solution: A computer-readable storage medium storing a computer program that can be loaded by a processor and executed any of the above-described gas self-closing valve control methods.

[0024] In summary, this application includes at least one of the following beneficial technical effects: 1. By acquiring room images and determining the target image corresponding to the gas equipment, the system collects the identity of personnel and their corresponding behaviors from the target image while the gas self-closing valve is open. The personnel identity includes their level of self-awareness. The opening degree of the gas self-closing valve is then adjusted according to the personnel identity and behavior. This makes the control of the gas self-closing valve no longer solely dependent on physical parameter thresholds such as pressure, flow rate, or concentration. Instead, it can make comprehensive decisions based on the cognitive characteristics and actual behavior of the people in the kitchen. This achieves a shift from indiscriminate passive shut-off to proactive differentiated protection based on personnel perception and behavioral understanding, effectively balancing gas safety and user experience. 2. By collecting the self-awareness level of personnel from the target image, the personnel in the kitchen are divided into first personnel and second personnel. Different opening adjustment strategies are implemented according to different personnel combination scenarios. This realizes the technical upgrade from a one-size-fits-all valve closing to graded protection based on personnel's cognitive ability. This avoids frequent accidental valve closing interference for personnel with self-awareness ability, and strengthens the proactive protection for personnel who lack the ability to identify danger. 3. When a first person and a second person are present in the kitchen at the same time, the initial control degree is obtained by acquiring the characteristics of both persons, and a correction coefficient is generated by combining the behavior and interaction status of the first person. Finally, the behavioral control degree is calculated in the form of a product, which realizes the dynamic quantitative assessment of the guardian's actual intervention ability. This allows the valve closing decision to truly reflect the degree of protection of the second person in the current scenario, reducing the false alarm rate and the false alarm rate. Attached Figure Description

[0025] Figure 1 This is a schematic flowchart of a gas self-closing valve control method disclosed in an embodiment of this application.

[0026] Figure 2 This is a flowchart illustrating a method for adjusting the opening of a gas self-closing valve as disclosed in an embodiment of this application.

[0027] Figure 3 This is a flowchart illustrating a method for calculating behavioral control degree disclosed in an embodiment of this application.

[0028] Figure 4 This is a flowchart illustrating a method for calculating a correction coefficient disclosed in an embodiment of this application.

[0029] Figure 5 This is a schematic flowchart of a valve control method for different second persons disclosed in an embodiment of this application.

[0030] Figure 6 This is a flowchart illustrating an early warning processing method for a first sub-category of personnel disclosed in an embodiment of this application.

[0031] Figure 7 This is a flowchart illustrating a method for handling early warnings for a second sub-category of personnel disclosed in an embodiment of this application.

[0032] Figure 8 This is a schematic diagram of a gas self-closing valve control system disclosed in an embodiment of this application. Detailed Implementation

[0033] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figures 1 to 8 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0034] This application discloses a method for controlling a gas self-closing valve. (Refer to...) Figure 1 The method includes: Step S101: Acquire room images.

[0035] Room images refer to digital images containing scenes inside a room, obtained through image acquisition devices. These devices can be cameras installed on the kitchen ceiling or walls, or camera modules built into smart gas stoves. Their coverage area includes at least the area where the gas appliance is located and the surrounding activity area.

[0036] For example, a camera can be installed in the corner of the ceiling of a family kitchen to capture real-time images of the entire kitchen, including the gas stove, range hood, countertop, and the activities of all people in the kitchen.

[0037] Step S102: Determine the target image from the room image based on the location of the gas appliance corresponding to the gas self-closing valve in the house.

[0038] The target image is a local image region cropped or extracted from the room image. The target image only contains the gas equipment itself and the space within a preset range around it, in order to reduce the interference of irrelevant background information on subsequent human behavior analysis.

[0039] For example, the specific coordinate range of the gas stove in the kitchen can be determined in advance through calibration. After the camera captures the overall room image, a sub-image containing only the gas stove, the area above the stove, and the area on the ground in front of the stove can be cut out from the original image based on the coordinate range. This sub-image is the target image.

[0040] Step S103: When the actual opening degree of the gas self-closing valve is greater than the preset opening degree, collect the personnel identity and the corresponding personnel behavior from the target image. The personnel identity includes the level of self-awareness.

[0041] Actual opening degree refers to the percentage of the gas self-closing valve core that is currently open, while preset opening degree is a pre-set opening threshold. When the actual opening degree is greater than the threshold, it indicates that the gas is in a usable flow state; when the actual opening degree is less than or equal to the threshold, it indicates that the gas has been cut off or is in a state of slight leakage. For example, the preset opening degree is 5%.

[0042] Self-awareness level refers to an individual's ability to recognize and judge the safety of their own behavior and its potential consequences. People with high self-awareness levels can proactively identify and avoid dangerous behaviors during gas usage, such as actively checking whether the gas is turned off and understanding the risks of gas leaks; people with low self-awareness levels lack this ability to identify dangers, such as children, some elderly people with cognitive impairments, or adults who are intoxicated.

[0043] Optionally, perform personnel detection on the target image to identify each human body object in the target image; for each human body object, extract the identity label of the human body object through a personnel identification model to obtain the personnel identity, the identity label including age category and self-awareness level score; for each human body object, extract the personnel behavior of the human body object through a personnel behavior identification model, the personnel behavior including behavior category and corresponding behavior parameters.

[0044] Step S104: Adjust the opening degree of the gas self-closing valve according to the personnel's identity and behavior.

[0045] For example, if the identified person is a child and the person touches a knob, the behavior is judged to be of high risk. Therefore, a closing command is sent to the drive circuit of the gas self-closing valve, causing the valve core of the gas self-closing valve to rotate to the 0% position, and the gas is cut off.

[0046] By adopting the above technical solution, by acquiring room images and determining the target image corresponding to the gas equipment, and with the gas self-closing valve in the open state, the identity of the personnel and their corresponding behaviors are collected from the target image. The personnel identity includes their level of self-awareness. Then, the opening degree of the gas self-closing valve is adjusted according to the personnel identity and behavior. This makes the control of the gas self-closing valve no longer solely rely on physical parameter thresholds such as pressure, flow rate, or concentration, but can make comprehensive decisions based on the cognitive characteristics and actual behavior of the people in the kitchen. This realizes the transformation from indiscriminate passive cut-off to proactive differentiated protection based on personnel perception and behavioral understanding, effectively balancing gas safety and user experience.

[0047] This application discloses a method for adjusting the opening degree of a gas self-closing valve. (Refer to...) Figure 2 The method includes: Step S201: Based on the self-awareness level in the personnel identity, divide the personnel in the personnel image into first personnel and second personnel. The self-awareness level of the first personnel is greater than a preset threshold, and the self-awareness level of the second personnel is less than or equal to the preset threshold.

[0048] The first person refers to someone whose level of self-awareness is higher than a preset threshold. For example, the first person is an adult with normal or high risk recognition ability who can independently and safely use gas appliances and has the ability to notice and stop dangerous behavior by other people.

[0049] The second person refers to someone whose self-awareness level is below or equal to a preset threshold. For example, the second person includes children, the elderly, people with cognitive impairment, or adults in a state of impaired consciousness. These people have difficulty identifying the dangers in the use of gas on their own and require external supervision.

[0050] The preset threshold is a critical value used to distinguish between high and low levels of self-awareness. For example, if the average self-awareness level of a normal adult is set at 70 points, the preset threshold is set to 70 points.

[0051] Step S202: If the personnel image only includes the first person, adjust the opening of the gas self-closing valve according to the personnel behavior and the first dangerous behavior.

[0052] The first dangerous act refers to an action or state performed by the first person who may cause a gas safety accident. Because the first person has a high level of self-awareness, the criteria for judging their dangerous behavior are relatively lenient, and the valve is only shut off when there is a clear violation. For example, first dangerous acts include leaving a burning gas stove for an extended period of time or using flammable spray near a gas stove.

[0053] Step S203: If the personnel image includes at least one first person and at least one second person, then generate the degree of control of the first person over the behavior of the second person. The degree of control is used to describe the degree of intervention in the behavior of other people.

[0054] In scenarios where both a first person and a second person are present, it cannot be simply assumed that the presence of a first person equates to safety. This is because the first person may not notice the dangerous behavior of the second person, or may be unable to intervene in time due to physical barriers (such as a stove). By quantifying the degree of behavioral control, the actual intervention capability of the first person can be incorporated into risk assessment, avoiding underestimation due to overestimating the effectiveness of supervision.

[0055] A higher level of behavioral control indicates that the first person is more capable of promptly detecting and stopping the dangerous actions of the second person; a lower level of behavioral control indicates that although the first person is present, they may be unable to intervene effectively due to distraction or physical obstruction. For example, if a 35-year-old man (first person) and a 6-year-old girl (second person) appear in the target image, it is necessary to assess the man's ability to supervise the girl and generate a behavioral control score, such as 0.6, indicating that the man has some intervention ability but certain limitations.

[0056] Step S204: Adjust the opening degree of the gas self-closing valve according to the degree of control over behavior and personnel behavior.

[0057] Optionally, obtain the weight of dangerous behavior corresponding to personnel behavior; according to the preset formula: (1-behavior control degree)×dangerous behavior weight, obtain the opening degree of the gas self-closing valve.

[0058] Step S205: If the personnel image only includes the second person, then adjust the opening degree of the gas self-closing valve according to the personnel behavior and the second dangerous behavior, where the danger level of the second dangerous behavior is greater than that of the first dangerous behavior.

[0059] Secondary hazardous behaviors refer to actions or states performed by a second person that have a higher probability of causing a gas accident, and their degree of danger is higher than that of primary hazardous behaviors. Because secondary persons lack the ability to recognize danger, even behaviors that may not seem serious to adults can pose significant risks; therefore, the criteria for judgment are more stringent. Examples of secondary hazardous behaviors include touching knobs or briefly leaving the area unattended.

[0060] By adopting the above technical solution, and by collecting the self-awareness level of personnel from the target image, the personnel in the kitchen are divided into first personnel and second personnel. Different opening adjustment strategies are implemented according to different personnel combination scenarios. This realizes the technical upgrade from a one-size-fits-all valve closure to graded protection based on personnel's cognitive ability. It avoids frequent accidental valve closure interference for personnel with self-awareness and strengthens the proactive protection for personnel who lack the ability to identify danger.

[0061] This application discloses a method for calculating behavioral control degree. (Refer to...) Figure 3 The method includes: Step S301: Based on the target image, obtain the first person features of the first person and the second person features of the second person.

[0062] The first person characteristics refer to the set of attributes used to characterize the first person's guardianship ability. The first person characteristics include, but are not limited to, age, height, body type, facial orientation, body posture, and relative position with the second person.

[0063] Secondary person characteristics refer to a set of attributes used to characterize the degree of guardianship needs of a secondary person. Secondary person characteristics include, but are not limited to, age, height, weight, activity level, frequency of historical dangerous behaviors, etc.

[0064] Step S302: Obtain the initial degree of control based on the first and second personnel characteristics.

[0065] For example, the first person's characteristics and the second person's characteristics are input into the control degree model to obtain the initial control degree.

[0066] Initial control reflects the potential for the first person to intervene in the second person's actions under ideal conditions.

[0067] Step S303: Obtain the first person's behavior and the interaction state between the first person and the second person from the personnel behavior.

[0068] First-person behavior refers to the sequence of actions performed by the first person, which may include head turning, eye movement, body movements, and verbal interaction.

[0069] Interactive state refers to the two-way behavioral relationship between the first person and the second person, such as whether the first person is looking at the second person, whether they issue a stop command, whether they reach out to stop them, and whether the second person responds to the first person's command.

[0070] Step S304: Generate correction coefficients based on the behavior and interaction status of the first person.

[0071] The initial control level is based on static characteristics and cannot reflect dynamic factors such as whether attention is focused or whether there are physical obstacles. Therefore, generating a correction coefficient can quantify these dynamic factors into a multiplier factor, which can be used to calibrate the initial control level in real time, so that the final behavioral control level can truly reflect the actual intervention capability at the moment.

[0072] Step S305: Calculate the product of the initial control degree and the correction coefficient to obtain the behavioral control degree.

[0073] By adopting the above technical solution, when a first person and a second person are present in the kitchen at the same time, the initial degree of control is obtained by acquiring the characteristics of both persons, and a correction coefficient is generated by combining the behavior and interaction status of the first person. Finally, the behavioral control degree is calculated in the form of a product, which realizes the dynamic quantitative assessment of the guardian's actual intervention ability. This allows the valve closing decision to truly reflect the degree of protection of the second person in the current scenario, reducing the false alarm rate and the missed alarm rate.

[0074] This application discloses a method for calculating a correction coefficient. (Refer to...) Figure 4 The method includes: Step S401: Extract the gaze behavior features of the first person towards the second person from the behavior of the first person. The gaze behavior features include the angle between the gaze direction of the first person and the gaze direction of the line connecting the first person and the second person.

[0075] Gaze behavior features are a set of parameters used to quantify the degree of visual attention a first person gives to a second person.

[0076] For example, through head pose estimation and eye tracking algorithms, the following can be calculated: the direction of the first person's face is vector A, the direction of the line connecting the first person to the second person is vector B, and the angle between the two vectors is 25°. Then, the direction vector A is the direction of the gaze, and 25° is the angle of the gaze.

[0077] Step S402: Extract the degree of attention the first person pays to the second person from the interaction status.

[0078] Attention level is a comprehensive indicator that reflects the intensity of the first person's emotional and attentional investment in the second person. It can be quantified by behaviors such as whether the first person continuously tracks the second person's movements or whether they show a tense posture when the second person makes a dangerous move.

[0079] Optionally, the proportion of the first person's gaze falling on the second person per unit time is calculated; a first attention score is determined in a preset first mapping database based on the aforementioned proportion, and the first attention score is positively correlated with the aforementioned proportion; a second attention score is generated based on whether the second person shows a tense posture when performing a dangerous action; the first attention score and the second attention score are weighted and calculated to obtain the degree of attention.

[0080] Step S403: Extract the physical distance information between the first person and the second person from the target image. The physical distance information is used to characterize whether there is an object between them that prevents the first person from interfering with the behavior of the second person.

[0081] In real-world scenarios, even if the first person notices the dangerous behavior of the second person, if there are physical obstacles that cannot be quickly crossed (such as a tall stove or island), the first person cannot reach out to intercept or approach to stop them in time. Therefore, extracting physical distance information can incorporate spatial accessibility into the assessment and avoid overestimating the actual intervention capability.

[0082] Physical distance information includes, but is not limited to, the height, width, and detour distance of obstacles.

[0083] Step S404: Determine the interval penalty coefficient based on the physical interval information.

[0084] Optionally, an interval penalty coefficient corresponding to the physical interval information can be determined in a preset database.

[0085] Step S405: Generate correction coefficients based on gaze behavior characteristics, attention level, and interval penalty coefficient.

[0086] Optionally, a correction coefficient can be generated by weighting the gaze behavior features, attention level, and interval penalty coefficient.

[0087] By adopting the above technical solution, when generating the correction coefficient, the angle of the first person's gaze, the degree of attention, and the physical distance between them are comprehensively extracted, and the physical distance is converted into an interval penalty coefficient and included in the calculation of the correction coefficient. This achieves a multi-dimensional integrated judgment of the guardian's attention allocation, attention intensity, and physical accessibility, effectively avoiding the problem of overestimation of intervention ability due to physical obstacles such as stoves and islands, and improving the accuracy of control assessment.

[0088] This application discloses a valve control method for different second personnel. (Refer to...) Figure 5 The method includes: Step S501: Identify the subdivided identity type of the second person. The subdivided identity type includes at least a first subdivided type and a second subdivided type. The first subdivided type corresponds to a child, and the second subdivided type corresponds to an elderly person.

[0089] Subcategories of identity types are further subcategories under the second major category of individuals, used to distinguish the behavioral risk characteristics of different vulnerable groups. For example, by using a facial age estimation model to identify the second individual, we can obtain the first subcategories and the second subcategories.

[0090] In real-world scenarios, the behavioral risk patterns of children and the elderly differ fundamentally: children primarily exhibit curiosity and exploration, while the elderly mainly demonstrate cognitive decline. Using the same set of judgment rules indiscriminately will lead to numerous missed or incorrect judgments. Therefore, identifying and categorizing specific identity types can make subsequent risk behavior judgment rules more targeted.

[0091] Step S502: When the second person belongs to the first sub-type, extract the relative spatial parameters between the second person and the gas equipment from the person's behavior. The relative spatial parameters include the real-time distance and the direction of movement.

[0092] Relative spatial parameters refer to the distance values ​​and movement direction vectors of key points on the second person's body relative to the reference point of the gas equipment. Key points on the body include hands, torso, etc., and reference points of the gas equipment include the center of the knob, the center of the stove surface, etc.

[0093] Step S503: If the real-time distance and direction of movement meet the first judgment condition, then the personnel behavior is judged as the second dangerous behavior.

[0094] The first judgment criterion is a predefined logical rule used to determine whether a child's behavior constitutes a danger. This criterion can be set by technicians according to actual needs. In one implementation, the child's real-time distance is less than 0.5 meters, the direction of movement is towards the gas appliance, and the distance is continuously decreasing.

[0095] Step S504: When the second person belongs to the second sub-type, extract the operation sequence features of the second person from the personnel behavior.

[0096] Operation timing characteristics refer to the complete time sequence pattern of a second person operating gas equipment, which includes the start time, duration, operation sequence, and operation interval of each operation.

[0097] Step S505: If the operation timing characteristics match the preset abnormal behavior pattern, the personnel behavior is judged as the second dangerous behavior.

[0098] Abnormal behavior patterns refer to a predefined sequence of typical unsafe operations for the elderly population. The specific content can be set by technicians according to actual needs. For example, abnormal behavior patterns include leaving immediately after turning on the stove for more than a threshold time, repeatedly turning the knob on and off without ever lighting the stove, and placing non-cooking items on the stove while it is running.

[0099] By adopting the above technical solution, in high-risk scenarios where only a second person is present, the system can further identify the second person's specific identity type, such as whether they are a child or an elderly person. For different specific types, the system uses judgment conditions based on spatial parameters and abnormal behavior pattern matching based on operation sequence characteristics. This achieves targeted adaptation to the behavioral characteristics of different vulnerable groups, enabling the security protection strategy to fit the actual risk performance of each group and improving the accuracy of dangerous behavior identification.

[0100] This application discloses an early warning processing method for a first sub-category of personnel. (Refer to...) Figure 6 The method includes: Step S601: Obtain the real-time distance and direction of movement of the second person.

[0101] Step S602: When the real-time distance has entered the preset high-risk zone, the direction of movement is towards the gas equipment, and the real-time distance continues to decrease within the preset time period, the person's behavior is determined to be a second dangerous behavior.

[0102] The high-risk zone refers to the critical range where the distance to the gas appliance is too close. For example, the high-risk zone is [0, 0.4].

[0103] When the real-time distance has entered the preset high-risk zone, the direction of movement is towards the gas equipment, and the real-time distance continues to decrease within the preset time period, it indicates that the child is actively approaching the gas equipment. To ensure the child's safety, the valve shut-off action needs to be triggered immediately.

[0104] Step S603: When the second dangerous behavior is not determined based on the real-time distance and direction of movement, check whether the knob of the gas equipment is in the open state and whether no effective ignition behavior is detected in the target image.

[0105] If a child does not approach the appliance but the knob is already turned on and there is no flame, it may indicate that the child has already turned the knob on, or that the gas appliance is leaking. This situation is also very dangerous and should be considered as a second layer of inspection if the distance requirement is not met.

[0106] Step S604: If yes, obtain the shortest distance between any part of the second person's body and the knob. When the shortest distance is continuously less than the preset first contact threshold, determine that the person's behavior belongs to the second dangerous behavior.

[0107] The first contact threshold is a preset empirical value used to determine whether a body part has already touched or is about to touch the knob. For example, the first contact threshold is 5 centimeters.

[0108] When the knob is on and there is no flame, the closer a child is to the knob, the higher the risk. This is because when the distance is less than the first contact threshold, the child can almost immediately operate the knob again. If the gas appliance is accidentally turned on, it may cause harm to the child. Therefore, it needs to be identified as a second dangerous behavior and valve shut-off measures should be taken.

[0109] Step S605: When the shortest distance does not determine the second dangerous behavior, extract the body center height and body center projection position of the second person from the target image.

[0110] The height of the body's center of gravity refers to the vertical distance between the center of gravity of the second person and the ground.

[0111] The projection position of the body's center of gravity refers to the two-dimensional coordinates of the vertical projection point of the body's center of gravity on the ground.

[0112] Step S606: When the height of the body's center of gravity exceeds the preset safe height threshold and the projected position of the body's center of gravity falls into the preset restricted area above the gas equipment, the person's behavior is determined to be a second dangerous behavior.

[0113] In addition to reaching out to touch, children may also climb on the stove, which can lead to falls, burns, or accidental activation of valves. Therefore, the height of the body's center of gravity and the position of its projection are key indicators for judging climbing behavior. When the distance judgment is not triggered, these posture parameters are needed to identify the climbing risk.

[0114] When a child's center of gravity rises above the safety threshold and their shadow falls on the stove, it means the child has climbed onto the stove, which is a very dangerous situation that could cause burns or gas leaks. The valve must be turned off immediately.

[0115] Step S607: When the body's center of gravity height and projected position are not determined to be a second dangerous behavior, check whether the flexible pipe connected to the gas equipment has a displacement exceeding the preset deformation.

[0116] Flexible pipe fittings refer to flexible metal or rubber hoses that connect gas equipment to fixed pipelines.

[0117] Children may pull on gas pipes, which could cause connections to loosen or detach, leading to gas leaks. This behavior is extremely dangerous, but it is difficult to detect effectively using distance or center of gravity parameters; therefore, specialized pipe displacement detection is required as a criterion for assessment.

[0118] Step S608: If yes, then determine whether there is a spatial correlation between the time of displacement and the limb movement trajectory of the second person.

[0119] Spatial correlation refers to whether the displacement event corresponds to the movement of a second person's limb in time and space. For example, the direction of the pipe displacement is consistent with the direction of the child's arm pulling, and the displacement time coincides with the time when the child's arm extends and retracts.

[0120] In real-world scenarios, pipe displacement can also be caused by external vibrations, not necessarily by children pulling on them. By determining the spatial correlation between displacement and a child's limb movements, false alarms can be eliminated, ensuring that a hazard assessment is only triggered when a child actively pulls on the pipe.

[0121] Step S609: When spatial correlation exists, determine that the person's behavior is a second dangerous behavior.

[0122] By adopting the above technical solution, for children as the second type of person, a progressive investigation is carried out from four aspects: real-time distance and direction of movement, contact distance between body parts and knobs when the knobs are turned on, height of body center of gravity and projection position, and abnormal displacement of flexible pipes. This covers a variety of common accidental behaviors of children, such as actively approaching and touching knobs, climbing stoves and pulling gas pipes, forming a complete behavioral risk protection mechanism and reducing the possibility of missed detection.

[0123] This application discloses an early warning processing method for a second sub-category of personnel. (Refer to...) Figure 7 The method includes: Step S701: After detecting that the gas equipment has been started, start the first preset time window timing; The first preset time window refers to the duration of the time counted from the start of the gas appliance. The length of the first preset time window can be set to 30 seconds.

[0124] Step S702: When a second person is detected leaving the monitoring area covered by the target image within the first preset time window, and the time of leaving continues to reach the second preset time window, it is determined that the person's behavior matches the abnormal behavior pattern.

[0125] The second preset time window refers to the shortest duration for which personnel continuously leave the monitored area.

[0126] When a second person is detected leaving the monitoring area covered by the target image within the first preset time window, and the time spent away reaches the second preset time window, it indicates that the elderly person has likely left the kitchen to do something else after turning on the gas stove and has not returned for a long time, which may cause the cookware to burn dry or the gas to burn empty. To ensure safety, the valve needs to be turned off in time.

[0127] Step S703: When the time of continuous departure is not matched, while the gas equipment is running, detect whether the relative stillness of the second person in the monitoring area continues for more than a third preset time window.

[0128] The third preset time window refers to the longest allowed period of relative stillness, for example, the third preset time window is set to 10 minutes.

[0129] Besides leaving the kitchen, elderly people may also appear in front of the stove but remain motionless for a long time, such as sitting in a daze, experiencing temporary incapacitation, or falling asleep. This can also cause the pot to burn dry, so it is necessary to turn off the valve in time.

[0130] Step S704: If yes, determine whether a predetermined action sequence related to cooking operation is identified from the target image within the third preset time window.

[0131] A pre-defined sequence of actions refers to a set of normal cooking operations, such as flipping the pan, adding seasonings, and checking the food in the pan.

[0132] Step S705: When no predetermined action sequence is identified, determine that the personnel behavior matches the abnormal behavior pattern.

[0133] After confirming that the elderly person has not performed the predetermined action sequence, it can be inferred that the elderly person is not cooking normally, but is in an abnormal state. At this time, it should be determined that the elderly person has an abnormality and valve shut-off measures should be taken.

[0134] Step S706: When a predetermined action sequence is identified, within a preset statistical period, count the number of times the second person performs the on-off cycle on the operating components of the gas equipment.

[0135] The number of on-off cycles refers to the total number of times the gas knob is turned from the off position to the on position and then back to the off position.

[0136] Step S707: When the number of on-off cycles exceeds the preset number and no corresponding valid ignition event is detected in the target image after each on-off operation, it is determined that the personnel behavior matches the abnormal behavior pattern.

[0137] When the number of on-off cycles exceeds the preset number, and no corresponding valid ignition event is detected in the target image after each on-off operation, it indicates that the elderly person is unaware that the gas appliance cannot be turned on and is in a state of cognitive confusion, requiring intervention by shutting off the valve.

[0138] Step S708: When the number of on-off cycles does not match, detect from the target image whether a second person has placed an item on a running gas device.

[0139] In real-world scenarios, elderly people may also place unsuitable items for heating, such as plastic bottles, rags, and paper products, on running gas appliances, which could directly cause a fire hazard.

[0140] Step S709: If yes, determine whether the item belongs to the preset cooking utensil category.

[0141] Cookware category refers to the collection of appliances heated on a gas stove, including metal or heat-resistant ceramic products such as woks, soup pots, steamers, and kettles.

[0142] Step S710: When the item does not belong to the cooking utensil category, determine that the personnel behavior matches the abnormal behavior pattern.

[0143] By adopting the above technical solution, for the second type of elderly people, the following behaviors were detected in sequence: the duration of time after the gas appliance is started and the person leaves the monitoring area; the duration of time in front of the stove where the person is relatively still and not cooking; the correspondence between the number of on-off cycles and effective ignition events; and the behavior of placing non-cooking items on the running gas appliance. This covered typical cognitive impairment behaviors such as forgetting to start the stove, inability to use the stove, repeated ineffective operations, and placement of abnormal items, thus achieving in-depth identification of gas usage risks for the elderly.

[0144] Based on the same inventive concept, this application provides a gas self-closing valve control system, please refer to... Figure 8 The system includes: The acquisition module 801 is used to acquire room images. Based on the location of the gas appliance corresponding to the gas self-closing valve in the house, the target image is determined from the room images.

[0145] The calculation module 802 is used to collect the identity of the person and the corresponding behavior of the person from the target image when the actual opening degree of the gas self-closing valve is greater than the preset opening degree. The identity of the person includes the level of self-awareness.

[0146] The processing module 803 is used to adjust the opening degree of the gas self-closing valve according to the personnel's identity and behavior.

[0147] By adopting the above technical solution, by acquiring room images and determining the target image corresponding to the gas equipment, and with the gas self-closing valve in the open state, the identity of the personnel and their corresponding behaviors are collected from the target image. The personnel identity includes their level of self-awareness. Then, the opening degree of the gas self-closing valve is adjusted according to the personnel identity and behavior. This makes the control of the gas self-closing valve no longer solely rely on physical parameter thresholds such as pressure, flow rate, or concentration, but can make comprehensive decisions based on the cognitive characteristics and actual behavior of the people in the kitchen. This realizes the transformation from indiscriminate passive cut-off to proactive differentiated protection based on personnel perception and behavioral understanding, effectively balancing gas safety and user experience.

[0148] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0149] This application provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as a gas self-closing valve control method.

[0150] Computer storage media include, for example, USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media that can store program code.

[0151] Based on the same inventive concept, embodiments of this application provide a smart terminal, including a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as a gas self-closing valve control method.

[0152] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0153] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A method for controlling a gas self-closing valve, characterized in that, include: Capture room images; The target image is determined from the room image based on the location of the gas appliance corresponding to the gas self-closing valve in the house; When the actual opening of the gas self-closing valve is greater than the preset opening, the personnel identity and the corresponding personnel behavior are collected from the target image, and the personnel identity includes the level of self-awareness. Adjust the opening degree of the gas self-closing valve according to the identity and behavior of the personnel.

2. The gas self-closing valve control method according to claim 1, characterized in that, Adjusting the opening of the gas self-closing valve based on the person's identity and behavior includes: Based on the level of self-awareness in the personnel identity, the personnel in the personnel image are divided into a first person and a second person. The self-awareness level of the first person is greater than a preset threshold, and the self-awareness level of the second person is less than or equal to the preset threshold. If the personnel image only includes the first personnel, then the opening degree of the gas self-closing valve is adjusted according to the personnel behavior and the first dangerous behavior; If the personnel image includes at least one first person and at least one second person, then the degree of control of the first person over the behavior of the second person is generated, the degree of control of the behavior is used to describe the degree of intervention in the behavior of other persons; the opening degree of the gas self-closing valve is adjusted according to the degree of control of the behavior and the behavior of the personnel. If the personnel image only includes the second person, then the opening degree of the gas self-closing valve is adjusted according to the personnel behavior and the second dangerous behavior, where the danger level of the second dangerous behavior is greater than the danger level of the first dangerous behavior.

3. The gas self-closing valve control method according to claim 2, characterized in that, The generation of the degree of control the first person has over the behavior of the second person includes: Based on the target image, obtain the first person features of the first person and the second person features of the second person; Based on the first personnel characteristics and the second personnel characteristics, the initial degree of control is obtained; From the personnel behaviors, obtain the first personnel behavior corresponding to the first personnel and the interaction state between the first personnel and the second personnel; Based on the behavior of the first person and the interaction state, a correction coefficient is generated; The behavioral control degree is obtained by multiplying the initial control degree by the correction coefficient.

4. The gas self-closing valve control method according to claim 3, characterized in that, The step of generating a correction coefficient based on the first person's behavior and the interaction state includes: Extract the gaze behavior features of the first person towards the second person from the behavior of the first person, the gaze behavior features including the angle between the gaze direction of the first person and the direction of the line connecting the first person and the second person; Extract the degree of attention the first person pays to the second person from the interaction state; Physical distance information between the first person and the second person is extracted from the target image. The physical distance information is used to characterize whether there is an object between them that prevents the first person from interfering with the behavior of the second person. Based on the physical interval information, determine the interval penalty coefficient; The correction coefficient is generated based on the gaze behavior characteristics, the level of attention, and the interval penalty coefficient.

5. The gas self-closing valve control method according to claim 2, characterized in that, The step of adjusting the opening of the gas self-closing valve based on the personnel behavior and the second dangerous behavior includes: Identify the subdivided identity type of the second person, the subdivided identity type including at least a first subdivided type and a second subdivided type, the first subdivided type corresponding to a child and the second subdivided type corresponding to an elderly person; When the second person belongs to the first sub-category, the relative spatial parameters between the second person and the gas equipment are extracted from the person's behavior. The relative spatial parameters include real-time distance and direction of movement. If the real-time distance and the direction of movement meet the first determination condition, the person's behavior is determined to be the second dangerous behavior. When the second person belongs to the second sub-type, the operation timing features of the second person are extracted from the person's behavior; if the operation timing features match the preset abnormal behavior pattern, the person's behavior is determined to be the second dangerous behavior.

6. The gas self-closing valve control method according to claim 5, characterized in that, The method further includes: Obtain the real-time distance and direction of movement of the second person; When the real-time distance has entered a preset high-risk zone, the direction of movement is toward the gas equipment, and the real-time distance continues to decrease within a preset time period, the person's behavior is determined to be the second dangerous behavior. When the real-time distance and the direction of movement are not determined to be a second dangerous behavior, the system detects whether the knob of the gas device is in the on state and whether no effective ignition behavior is detected in the target image. If so, the shortest distance between any part of the second person's body and the knob is obtained. When the shortest distance is continuously less than a preset first contact threshold, the person's behavior is determined to be the second dangerous behavior. When the shortest distance is not determined to be a second dangerous behavior, the body center height and body center projection position of the second person are extracted from the target image; When the height of the body's center of gravity exceeds a preset safe height threshold, and the projected position of the body's center of gravity falls into a preset restricted area above the gas equipment, the person's behavior is determined to be the second dangerous behavior. When the body's center of gravity height and the body's center of gravity projection position are not determined to be a second dangerous behavior, the system detects whether the flexible pipe connected to the gas equipment has undergone displacement exceeding a preset deformation. If so, determine whether there is a spatial correlation between the time of the displacement and the limb movement trajectory of the second person; When the spatial correlation exists, the person's behavior is determined to be the second dangerous behavior.

7. The gas self-closing valve control method according to claim 5, characterized in that, The method further includes: After detecting that the gas equipment has been started, the first preset time window is started for timing; When the second person is detected leaving the monitoring area covered by the target image within the first preset time window, and the time spent leaving reaches the second preset time window, it is determined that the person's behavior matches the abnormal behavior pattern. When the duration of the departure does not match, while the gas equipment remains running, it is detected whether the relative stillness of the second person in the monitoring area lasts for more than a third preset time window. If so, determine whether a predetermined action sequence related to cooking operations has been identified from the target image within the third preset time window; When the predetermined action sequence is not identified, it is determined that the person's behavior matches the abnormal behavior pattern; When the predetermined action sequence is detected, the number of times the second person performs the on-off cycle on the operating components of the gas equipment is counted within a preset statistical period; When the number of on-off cycles exceeds a preset number, and no corresponding valid ignition event is detected in the target image after each on-off operation, the personnel behavior is determined to match the abnormal behavior pattern. When the number of on-off cycles does not match, detect from the target image whether the second person has placed an item on a running gas appliance; If so, determine whether the item belongs to a preset cooking utensil category; When the item does not belong to the cooking utensil category, the person's behavior is determined to match the abnormal behavior pattern.

8. A gas self-closing valve control system, characterized in that, The system is used to execute the gas self-closing valve control method as described in any one of claims 1 to 7, including: The acquisition module is used to acquire room images; and to determine the target image from the room images based on the location of the gas appliance corresponding to the gas self-closing valve in the house. The calculation module is used to collect the identity of the person and the corresponding behavior of the person from the target image when the actual opening degree of the gas self-closing valve is greater than the preset opening degree. The identity of the person includes the level of self-awareness. The processing module is used to adjust the opening degree of the gas self-closing valve according to the identity and behavior of the personnel.

9. A smart terminal, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and execute the method as described in any one of claims 1 to 7.

Images