A method and system for detecting a kitchen heat source

Abstract

The embodiment of the application discloses a kind of detection method and system of kitchen heat source, the method comprises: current kitchen environment information and the state information of target heat source are collected;According to current kitchen environment information, determine dynamic threshold, dynamic threshold is characterized the threshold of temperature corresponding to current kitchen environment information;According to the state information of target heat source, determine the attribute of target heat source;According to dynamic threshold, the attribute of target heat source, risk assessment is carried out, and risk assessment result is obtained;Intervention treatment is carried out based on risk assessment result.The detection method of kitchen heat source provided in the application can effectively reduce the false alarm rate by determining the dynamic threshold and judging the heat source attribute, and finally performing risk assessment, according to the risk assessment result, intervention treatment, can effectively reduce the false alarm rate, eliminate environmental interference while shortening the response delay, improve the accuracy and efficiency of detection.

Description

Technical Field

[0001] This invention relates to the field of kitchen safety monitoring technology, and in particular to a method and system for detecting kitchen heat sources. Background Technology

[0002] Fixed heat sources in the kitchen, such as gas stoves, induction cookers, and ovens, are core equipment in both home and commercial kitchens. Their safety is directly related to the personal safety and property safety of users, as well as the normal conduct of daily cooking activities.

[0003] To ensure the safety of fixed heat sources in the kitchen, various detection methods are currently employed, including those related to gas, electricity, temperature, and smoke. Combustible gas detectors are installed near gas appliances to monitor the concentration of combustible gas in the air in real time, automatically alarming and cutting off the gas supply when levels exceed safe limits. Infrared thermal imagers are used to detect the temperature of electrical wiring, sockets, and equipment surfaces to identify overheating hazards and prevent electrical fires. Smoke detectors are installed on the kitchen ceiling to detect smoke promptly and trigger alarms, suitable for early fire warnings. However, existing monitoring methods still have shortcomings. Traditional smoke detectors lack the ability to analyze heat source attributes, resulting in a high false alarm rate. They are also susceptible to environmental interference; strong direct sunlight or open ventilation significantly increases the probability of false triggering by traditional infrared sensors, causing users to actively disable the alarm function and increasing safety hazards. Traditional systems rely on centralized cloud processing, resulting in an average delay of several seconds from the occurrence of a fire to the user receiving the alarm, leading to low detection efficiency.

[0004] Therefore, it is particularly important to develop a method and system for detecting kitchen heat sources that can effectively reduce false alarm rates, shorten response delays, and improve detection accuracy and efficiency by determining dynamic thresholds, judging heat source attributes, conducting risk assessments, and intervening based on the risk assessment results. Summary of the Invention

[0005] To address the aforementioned technical problems, this application provides a method and system for detecting kitchen heat sources. By determining dynamic thresholds and judging heat source attributes, and finally conducting a risk assessment, intervention is carried out based on the risk assessment results. This addresses the current lack of a method and system for detecting kitchen heat sources that can effectively reduce false alarm rates, shorten response delays, and improve detection accuracy and efficiency.

[0006] The technical solution provided in this application is as follows: On the one hand, this application provides a method for detecting a kitchen heat source, the method comprising: Collect current kitchen environment information and target heat source status information; Based on the current kitchen environment information, a dynamic threshold is determined, wherein the dynamic threshold represents the temperature threshold corresponding to the current kitchen environment information; Based on the state information of the target heat source, determine the attributes of the target heat source; Based on the dynamic threshold and the properties of the target heat source, a risk assessment is performed to obtain the risk assessment result; Intervention measures will be taken based on the risk assessment results.

[0007] In some optional implementations, the current kitchen environment information includes the rate of temperature change, the rate of humidity change, and the solar radiation intensity. Determining the dynamic threshold based on the current kitchen environment information includes: Obtain the preset baseline threshold, the current first environmental compensation coefficient, and the second environmental compensation coefficient; The product of the temperature change rate and the humidity change rate is determined as the environmental change rate; The sum of the solar radiation intensity and the second environmental compensation coefficient is determined as the radiation intensity compensation value; The ratio of the environmental change rate to the radiation intensity compensation value is determined as the environmental impact factor. The product of the current first environmental compensation coefficient and the environmental impact factor is determined as the threshold adjustment range; The product of the baseline threshold and the threshold adjustment range is determined as the threshold adjustment value; The sum of the baseline threshold and the threshold adjustment value is determined as the dynamic threshold.

[0008] In some optional implementations, the state information of the target heat source includes infrared image frequency domain features, displacement information of the hot spot centroid, absorbed heat, and temperature change. Determining the attributes of the target heat source based on its state information includes: Obtain a first target model; the first target model is obtained by training a first preset model to predict the flicker state based on the frequency domain features of historical infrared images and the flicker state labels corresponding to the frequency domain features of the historical infrared images. The infrared image frequency domain features are input into the first target model to obtain the flashing state corresponding to the infrared image frequency domain features; Based on the displacement information of the hot spot centroid, the trajectory deviation of the hot spot centroid is determined; The specific heat capacity of the target heat source is determined based on the absorbed heat and the temperature change. The properties of the target heat source are determined based on the flickering state, the trajectory deviation of the hot spot centroid, and the specific heat capacity of the target heat source.

[0009] In some optional implementations, the hot spot centroid displacement information includes horizontal displacement and vertical displacement, and determining the trajectory deviation of the hot spot centroid based on the hot spot centroid displacement information includes: Obtain the preset displacement velocity threshold; The ratio of the product of the horizontal displacement and the vertical displacement to the displacement velocity threshold is determined as the attribute influence value; The trajectory deviation of the hot spot centroid is obtained by summing the influence values ​​of the attribute corresponding to several time points.

[0010] In some optional implementations, determining the specific heat capacity of the target heat source based on the absorbed heat and the temperature change includes: Obtain first preset information, which characterizes the relationship between the heat source and its mass; Based on the first preset information, the target heat source mass corresponding to the target heat source is obtained; The ratio of the absorbed heat to the product of the temperature change and the mass of the target heat source is determined as the specific heat capacity of the target heat source.

[0011] In some optional implementations, determining the properties of the target heat source based on the flicker state, the trajectory deviation of the hot spot centroid, and the specific heat capacity of the target heat source includes: Under the condition that the first preset condition is met, the target heat source is determined to be a fire heat source. The first preset condition is that the flashing state is a high-frequency flashing state, and the trajectory deviation of the hot spot centroid is greater than or equal to a preset trajectory deviation threshold, and the specific heat capacity of the target heat source is greater than or equal to a preset specific heat capacity threshold. Under the condition that the second preset condition is met, the target heat source is determined to be a cooking heat source. The second preset condition is that the flickering state is stable, the trajectory deviation of the hot spot centroid is less than the preset trajectory deviation threshold, and the specific heat capacity of the target heat source is less than the preset specific heat capacity threshold.

[0012] In some optional implementations, the step of performing a risk assessment based on the dynamic threshold and the properties of the target heat source to obtain a risk assessment result includes: Obtain the second target model; the second target model is obtained by risk assessment prediction training of the second preset model based on historical dynamic thresholds, attributes of historical heat sources and risk assessment labels corresponding to the historical heat sources; The dynamic threshold and the attributes of the target heat source are input into the second target model to obtain the risk assessment result of the target heat source.

[0013] In some optional implementations, the intervention based on the risk assessment results includes: If the risk assessment result is less than the first risk threshold, an alarm will be pushed to the system. If the risk assessment result is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, an audible and visual alarm is triggered, and the exhaust equipment in the kitchen is activated. If the risk assessment result is greater than the second risk threshold, a telephone alarm will be triggered, and the gas appliances in the kitchen will be shut off.

[0014] In some optional implementations, the current kitchen environment information further includes the current wind speed and the current solar altitude angle. After determining the dynamic threshold based on the current kitchen environment information, the method further includes: When the solar radiation intensity is greater than a preset radiation intensity threshold, a target radiation intensity reference value corresponding to the current solar altitude angle is obtained based on second preset information, wherein the second preset information characterizes the correspondence between the solar altitude angle and the radiation intensity reference value. If the difference between the target radiation intensity reference value and the solar radiation intensity is greater than a preset threshold, the solar radiation intensity is updated to the target radiation intensity reference value, and the dynamic threshold is redefined. When the wind speed is greater than a preset wind speed threshold, a target first environmental compensation coefficient corresponding to the current wind speed is obtained based on third preset information, wherein the third preset information represents the correspondence between wind speed and the first environmental compensation coefficient. Update the current first environmental compensation coefficient to the target first environmental compensation coefficient, and redetermine the dynamic threshold.

[0015] On the other hand, this application provides a kitchen heat source detection system, which includes a data acquisition module, an edge computing module, and an execution module; The data acquisition module is used to collect current kitchen environment information and target heat source status information; The edge computing module is used to determine a dynamic threshold based on the current kitchen environment information; and to determine the attributes of the target heat source based on the status information of the target heat source; and to perform a risk assessment based on the dynamic threshold and the attributes of the target heat source to obtain a risk assessment result. The execution module is used to perform intervention based on the risk assessment results.

[0016] On the other hand, this application provides an electronic device including a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement the kitchen heat source detection method as described in any of the above embodiments.

[0017] On the other hand, this application provides a computer-readable storage medium storing at least one instruction or at least one program, which is loaded and executed by a processor to implement the kitchen heat source detection method as described in any of the above embodiments.

[0018] The kitchen heat source detection method provided in this application includes: collecting current kitchen environment information and target heat source status information; determining a dynamic threshold based on the current kitchen environment information, wherein the dynamic threshold represents a temperature threshold corresponding to the current kitchen environment information; determining the attribute of the target heat source based on the target heat source status information; performing a risk assessment based on the dynamic threshold and the target heat source attribute to obtain a risk assessment result; and performing intervention processing based on the risk assessment result. By determining the dynamic threshold and judging the heat source attribute, and finally performing a risk assessment and intervention processing based on the risk assessment result, the false alarm rate can be effectively reduced, environmental interference can be eliminated, and the accuracy and efficiency of detection can be improved. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of a kitchen heat source detection method according to an embodiment of the present invention; Figure 2 This is a structural diagram of a kitchen heat source detection system according to an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0022] The term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of this application. In the description of this application, it should be understood that the terms "upper," "lower," "top," "bottom," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. Moreover, the terms "first," "second," etc., are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein.

[0023] When a numerical range is disclosed herein, the range is considered continuous and includes the minimum and maximum values ​​of the range, as well as every value between the minimum and maximum values. Furthermore, when the range refers to an integer, it includes every integer between the minimum and maximum values ​​of the range. Additionally, when multiple ranges are provided to describe a feature or characteristic, the ranges may be combined. In other words, unless otherwise specified, all ranges disclosed herein should be understood to include any and all subranges to which they are included. For example, a specified range from “1 to 10” should be considered to include any and all subranges between the minimum value 1 and the maximum value 10. Exemplary subranges of the range 1 to 10 include, but are not limited to, 1 to 6.1, 3.5 to 7.8, 5.5 to 10, etc.

[0024] Current detection methods lack heat source attribute analysis capabilities, resulting in high false alarm rates. Environmental interference further increases the probability of false triggering, leading users to proactively disable alarm functions. Furthermore, reliance on centralized cloud processing results in low detection efficiency. Therefore, to effectively reduce false alarm rates, eliminate environmental interference, shorten response delays, and improve the accuracy and efficiency of kitchen heat source detection, this application provides a method and system for detecting kitchen heat sources.

[0025] Please see Figure 1 , Figure 1 This is a flowchart of a kitchen heat source detection method according to an embodiment of the present invention. In one aspect, this application provides a kitchen heat source detection method, which includes: S101. Collect current kitchen environment information and target heat source status information.

[0026] Optionally, a sensor array and optical components can be installed to collect information on the current kitchen environment and the status of the target heat source. The infrared thermal imager can be installed at a height of 1.8-2.2 meters with a downward angle of 30°; the visible light camera needs to cover the stove area and avoid being blocked by the range hood; the temperature and humidity sensor is installed in a corner of the kitchen, away from the stove and sink.

[0027] Optionally, the sensor array includes an infrared thermal imaging sensor, a visible light camera, a temperature and humidity sensor, a barometric pressure sensor, and a millimeter-wave radar; the optical components include an electrically adjustable infrared filter and a self-cleaning coated lens. Infrared thermal imaging is used to locate the heat source and its temperature distribution; the visible light camera is used to identify the shape and trajectory of the heat source; the temperature and humidity sensor is used to compensate for environmental interference and eliminate the influence of water vapor and oil fumes on the infrared sensor; and the millimeter-wave radar is used to measure the three-dimensional velocity of the heat source and detect the direction and speed of the flame jet.

[0028] By collecting the above multi-dimensional information, the system can fully grasp the kitchen environment and heat source status, laying a data foundation for subsequent dynamic threshold calculation, heat source attribute identification and risk assessment.

[0029] S102. Based on the current kitchen environment information, determine a dynamic threshold, wherein the dynamic threshold represents the temperature threshold corresponding to the current kitchen environment information.

[0030] In an optional embodiment, the current kitchen environment information includes the rate of temperature change, the rate of humidity change, and the solar radiation intensity. Determining the dynamic threshold based on the current kitchen environment information includes: Obtain the preset baseline threshold, the current first environmental compensation coefficient, and the second environmental compensation coefficient; The product of the temperature change rate and the humidity change rate is determined as the environmental change rate; The sum of the solar radiation intensity and the second environmental compensation coefficient is determined as the radiation intensity compensation value; The ratio of the environmental change rate to the radiation intensity compensation value is determined as the environmental impact factor. The product of the current first environmental compensation coefficient and the environmental impact factor is determined as the threshold adjustment range; The product of the baseline threshold and the threshold adjustment range is determined as the threshold adjustment value; The sum of the baseline threshold and the threshold adjustment value is determined as the dynamic threshold.

[0031] Optionally, the temperature change rate refers to the range of temperature change in the kitchen environment per unit time, which is collected by a temperature and humidity sensor; the humidity change rate refers to the range of humidity change in the kitchen environment per unit time, which is also collected by a temperature and humidity sensor, and is used to reflect the influence of water vapor and other substances in the environment on temperature monitoring; the solar radiation intensity refers to the intensity of radiation energy from sunlight shining into the kitchen, which is collected by a light intensity sensor.

[0032] Optionally, the preset baseline threshold is an initial temperature threshold pre-set by the system, which serves as the basic reference value for dynamic threshold calculation. Its value is set based on the safe temperature range of common heat sources in the kitchen and can be initially configured according to different kitchen scenarios.

[0033] Optionally, the first environmental compensation coefficient is a parameter used to adjust the degree of influence of environmental impact factors on the threshold. It reflects the weight of temperature and humidity on the overall threshold and can be dynamically optimized according to the environmental characteristics of the kitchen during long-term use. For example, in a humid environment, its value may be appropriately increased to enhance sensitivity to humidity changes. The second environmental compensation coefficient is used to compensate for and correct solar radiation intensity, avoiding abnormal radiation intensity compensation values ​​due to excessive solar radiation. Its value is a preset constant, and its main function is to balance the proportion of solar radiation intensity in the calculation and reduce the interference of extreme radiation values ​​on environmental impact factors.

[0034] Optionally, the environmental change rate is the product of the temperature change rate and the humidity change rate, which comprehensively reflects the overall trend of temperature and humidity changes in the kitchen environment. For example, when frying or stir-frying, the temperature rises rapidly and the humidity changes due to the increase in oil fumes, at which time the environmental change rate will increase significantly.

[0035] Optionally, the radiation intensity compensation value is the sum of the solar radiation intensity and the second environmental compensation coefficient, which serves to make a basic correction to the solar radiation intensity.

[0036] Optionally, the environmental impact factor is the ratio of the rate of environmental change to the radiation intensity compensation value. The environmental impact factor quantifies the proportion of current environmental changes relative to the impact of solar radiation. For example, when ventilation is good, the rate of temperature change decreases, and the environmental impact factor will decrease accordingly.

[0037] Optionally, the threshold adjustment range is determined by the product of the current first environmental compensation coefficient and the environmental impact factor, reflecting the adjustment strength of environmental factors on the benchmark threshold. The larger the first environmental compensation coefficient, the more significant the adjustment effect of environmental impact factors on the threshold.

[0038] Optionally, the final dynamic threshold will be adjusted based on changes in temperature, humidity, and solar radiation. When sunlight is direct, radiation intensity increases, the radiation intensity compensation value rises, and the environmental impact factor decreases, making the dynamic threshold closer to the baseline threshold. Conversely, when temperature and humidity change rapidly in the kitchen, the environmental impact factor increases, and the dynamic threshold will be significantly higher than the baseline threshold, thus increasing sensitivity to high-risk heat sources. The formula for calculating the dynamic threshold is: ; Where T is the calculated dynamic threshold, representing the threshold for the measured temperature corresponding to the current kitchen environment information. The preset baseline threshold, For the rate of temperature change, Let I be the rate of change of humidity, and I be the intensity of solar radiation. The first environmental compensation coefficient, This is the second environmental compensation coefficient.

[0039] The calculation of dynamic thresholds integrates environmental factors such as temperature and humidity changes and solar radiation, enabling real-time adaptive adjustment of the threshold. This solves the problem of high false alarm rates in traditional fixed thresholds under environmental interference. When direct sunlight causes a sudden increase in solar radiation intensity, the radiation intensity compensation value increases, the environmental impact factor decreases, and the dynamic threshold does not rise excessively, avoiding misjudgment as a high-temperature risk. When the temperature and humidity in the kitchen change rapidly due to cooking, the environmental impact factor increases, and the dynamic threshold rises, ensuring accurate monitoring of real high-temperature heat sources. This mechanism enables the system to dynamically optimize judgment criteria based on the real-time environment, balancing detection sensitivity and anti-interference capability.

[0040] In an optional embodiment, the current kitchen environment information further includes the current wind speed and the current solar altitude angle. After determining the dynamic threshold based on the current kitchen environment information, the method further includes: When the solar radiation intensity is greater than a preset radiation intensity threshold, a target radiation intensity reference value corresponding to the current solar altitude angle is obtained based on second preset information, wherein the second preset information characterizes the correspondence between the solar altitude angle and the radiation intensity reference value. If the difference between the target radiation intensity reference value and the solar radiation intensity is greater than a preset threshold, the solar radiation intensity is updated to the target radiation intensity reference value, and the dynamic threshold is redefined. When the wind speed is greater than a preset wind speed threshold, a target first environmental compensation coefficient corresponding to the current wind speed is obtained based on third preset information, wherein the third preset information represents the correspondence between wind speed and the first environmental compensation coefficient. Update the current first environmental compensation coefficient to the target first environmental compensation coefficient, and redetermine the dynamic threshold.

[0041] Optionally, a preset radiation intensity threshold is used to determine whether solar radiation is strong enough to potentially interfere with temperature monitoring. The preset radiation intensity threshold can be set according to actual business needs; for example, the preset radiation intensity threshold is 500W / m². 2 Solar radiation intensity greater than 500W / m 2 At that time, the solar radiation correction mechanism is activated, and the system switches the infrared filter to narrowband mode to reduce interference.

[0042] Optionally, the current solar elevation angle is collected by optical components or associated sensors to establish a lookup model of solar elevation angle and radiation intensity. When the solar radiation intensity is abnormal, the solar radiation intensity value is dynamically corrected.

[0043] Optionally, solar radiation intensity is directly related to the solar altitude angle, and under normal circumstances, the two should match. The second preset information is a lookup model trained using historical data. The higher the solar altitude angle, the larger the corresponding radiation intensity reference value; the lower the solar altitude angle, the smaller the corresponding radiation intensity reference value. For example, at noon when the solar altitude angle is 80°, the radiation intensity reference value is approximately 800-1000 W / m². 2 .

[0044] Optionally, if the difference between the target radiation intensity reference value and the currently measured solar radiation intensity is greater than a preset threshold, it indicates that the currently measured solar radiation intensity is abnormal and needs to be corrected. Specifically, the current solar radiation intensity is updated to the target radiation intensity reference value, that is, the abnormal measurement value is replaced by the reasonable value predicted by the model, and the dynamic threshold is recalculated based on the updated solar radiation intensity.

[0045] Optionally, the current wind speed is collected by millimeter-wave radar, and the calculation formula is as follows: ; in, The current wind speed, is the phase difference, f is the Doppler frequency shift, and c is the electromagnetic wave propagation speed. It is used to evaluate the disturbance of airflow to ambient temperature. When the current wind speed is greater than 1 m / s, the system activates the anti-interference mode.

[0046] Optionally, the preset wind speed threshold can be set according to actual business needs. For example, the preset wind speed threshold can be 1 m / s. When the interference is active, the anti-interference mode is activated. This threshold is used to determine whether the airflow is strong enough to potentially interfere with the measurement of the rate of change of temperature and humidity.

[0047] Optionally, the third preset information is a correspondence table determined through experimental data. The higher the wind speed, the stronger the interference of airflow on temperature and humidity changes, requiring a reduction in the weight of the environmental change rate in the dynamic threshold calculation. The currently used first environmental compensation coefficient is updated to the target first environmental compensation coefficient, and the dynamic threshold is recalculated based on the updated coefficient. The first environmental compensation coefficient directly affects the threshold adjustment range. Reducing the value of the first environmental compensation coefficient will reduce the impact of the environmental change rate on the dynamic threshold, thereby avoiding interference with threshold judgment due to temperature change rate distortion caused by airflow.

[0048] By correcting for solar radiation intensity and adjusting the compensation coefficient based on wind speed, the interference of sunlight and airflow on threshold calculation is further eliminated, ensuring that the dynamic threshold can truly reflect the actual risk of kitchen heat sources. This ultimately reduces the false alarm rate, avoids dynamic threshold distortion caused by direct sunlight or strong winds, and improves the accuracy of the system's assessment of heat source risks.

[0049] S103. Determine the attributes of the target heat source based on the state information of the target heat source.

[0050] In an optional embodiment, the state information of the target heat source includes infrared image frequency domain features, displacement information of the hot spot centroid, absorbed heat, and temperature change. Determining the attributes of the target heat source based on its state information includes: Obtain a first target model; the first target model is obtained by training a first preset model to predict the flicker state based on the frequency domain features of historical infrared images and the flicker state labels corresponding to the frequency domain features of the historical infrared images. The infrared image frequency domain features are input into the first target model to obtain the flashing state corresponding to the infrared image frequency domain features; Based on the displacement information of the hot spot centroid, the trajectory deviation of the hot spot centroid is determined; The specific heat capacity of the target heat source is determined based on the absorbed heat and the temperature change. The properties of the target heat source are determined based on the flickering state, the trajectory deviation of the hot spot centroid, and the specific heat capacity of the target heat source.

[0051] Optionally, the frequency domain features of the infrared image are acquired by an infrared thermal imager. By performing frequency domain analysis on the infrared image, the flickering features of the heat source, such as high-frequency flickering or a stable state, are extracted to provide a basis for determining whether the heat source is a flame.

[0052] Optionally, the first target model is a machine learning model used to predict the flicker state corresponding to the frequency domain features of infrared images. Frequency domain features are those extracted after processing the infrared image using Fourier transform and other methods, reflecting the frequency variation of the heat source, such as the high-frequency fluctuations of a flame or the stable frequency characteristics of a fixed heat source. Flicker state labels are divided into high-frequency flicker and stable states, and these labels are either manually labeled or automatically generated from experimental data. Historical infrared image frequency domain features are input into the first preset model, with the corresponding flicker state labels as the supervised target. The model parameters are adjusted using algorithms such as backpropagation, enabling the model to learn the mapping relationship between frequency domain features and flicker states. After training, the first target model, which can be directly used for prediction, is obtained.

[0053] Optionally, an infrared image of the target heat source is acquired using an infrared thermal imager. The image is then converted to the frequency domain to extract real-time frequency domain features. These real-time infrared image frequency domain features are input into a first target model. Based on the mapping relationship learned during training, the model outputs the flickering state corresponding to the feature. For example, when a pan of oil catches fire, the infrared image frequency domain features of the flame exhibit high-frequency fluctuations, and the model will output "high-frequency flickering"; while when a gas stove is heating a pan normally, the infrared image frequency domain features are stable, and the model will output "stable state".

[0054] By accurately distinguishing the flashing state using the first target model, it is possible to preliminarily determine whether the heat source is a fire, providing an important basis for subsequent risk assessment, and ultimately improving the system's ability to distinguish between fire and normal cooking, thus reducing the false alarm rate.

[0055] In an optional embodiment, the hot spot centroid displacement information includes horizontal displacement and vertical displacement, and determining the trajectory deviation of the hot spot centroid based on the hot spot centroid displacement information includes: Obtain the preset displacement velocity threshold; The ratio of the product of the horizontal displacement and the vertical displacement to the displacement velocity threshold is determined as the attribute influence value; The trajectory deviation of the hot spot centroid is obtained by summing the influence values ​​of the attribute corresponding to several time points.

[0056] Optionally, the displacement information of the hot spot centroid is jointly acquired by an infrared thermal imager and a visible light camera. The displacement information of the hot spot centroid includes horizontal displacement and vertical displacement, which represent the positional changes of the hot spot centroid in the horizontal and vertical directions, respectively. Both are used together to calculate the trajectory deviation of the hot spot centroid, reflecting the motion stability of the heat source, such as the irregular movement of the flame or the stable state of a fixed cooking appliance.

[0057] Optionally, the preset displacement velocity threshold is a critical velocity value pre-set by the system, which is a critical standard characterizing the displacement of the hot spot centroid, and is determined based on the theoretical limit velocity of flame propagation or the normal displacement range of the cooking heat source.

[0058] Optionally, the formula for calculating the trajectory deviation of the hot spot centroid is: ; in, The deviation of the hot spot centroid from its trajectory. Let be the horizontal displacement at time t. Let be the vertical displacement at time t. The preset displacement velocity threshold is used. For each time point t, the attribute influence value is determined by the ratio of the product of the horizontal displacement and the vertical displacement to the displacement velocity threshold. The product of the horizontal and vertical displacements reflects the overall displacement amplitude of the hot spot centroid in the two-dimensional plane. The larger the product, the more significant the displacement. After dividing by the displacement velocity threshold, the displacement amplitude can be standardized, which facilitates the comparison of values ​​at different time points. The attribute influence values ​​of multiple consecutive time points (t=1 to t=n) are summed to obtain the trajectory deviation of the hot spot centroid. The trajectory deviation reflects the cumulative effect of the hot spot centroid displacement over a period of time. If the heat source is stable, the displacement at each time point is minimal. If it is a flame, the displacement is frequent and the amplitude is large, and the trajectory deviation will increase significantly.

[0059] Trajectory deviation is one of the key indicators for judging the attributes of a heat source. By calculating trajectory deviation, the motion stability of the heat source can be quantified. When the trajectory deviation is less than a preset threshold, it indicates that the displacement of the hot spot centroid is stable, which is likely a cooking heat source such as a pot or a gas stove with stable combustion. When the trajectory deviation is greater than or equal to the preset threshold, it indicates that the displacement of the hot spot centroid is irregular and has a large amplitude, which may be a fire heat source. This process effectively distinguishes the motion characteristics of cooking and unstable fire, provides an important basis for subsequent risk assessment, and reduces alarm deviations caused by misjudgment of the heat source motion state.

[0060] In an optional embodiment, determining the specific heat capacity of the target heat source based on the absorbed heat and the temperature change includes: Obtain first preset information, which characterizes the relationship between the heat source and its mass; Based on the first preset information, the target heat source mass corresponding to the target heat source is obtained; The ratio of the absorbed heat to the product of the temperature change and the mass of the target heat source is determined as the specific heat capacity of the target heat source.

[0061] Optionally, the heat absorption is monitored by an infrared thermal imager to detect the heat absorbed by the heat source over a certain period of time, and the specific heat capacity is calculated in combination with the temperature change of the heat source.

[0062] Optionally, the temperature change refers to the range of temperature change of the target heat source itself over a certain period of time, which is directly collected by an infrared thermal imager and used to help determine the activity level of the heat source.

[0063] Optionally, the first preset information is a pre-established correspondence between heat sources and their masses. It is a database or mapping table containing different types of heat sources and their corresponding masses, including the masses of common cooking heat sources and the equivalent masses of typical fire-related heat sources such as burning oil. This information is obtained through historical data statistics or experimental measurements and is used to quickly match the mass of the target heat source.

[0064] Optionally, the formula for calculating the specific heat capacity of the target heat source is: ; in, The specific heat capacity of the target heat source. To absorb heat, Let m be the temperature change and m be the target heat source mass. Specific heat capacity is one of the key indicators for distinguishing the properties of a heat source. Cooking heat sources, such as cookware, are usually made of metal, have a small specific heat capacity, and their temperature changes are relatively stable. The equivalent specific heat capacity of a fire heat source is large. Because energy is released quickly during combustion, the temperature rises sharply, and the estimated mass value is small, combined with characteristics such as high-frequency flickering and large trajectory deviation, it can be clearly identified as a fire.

[0065] By calculating specific heat capacity, the system can distinguish between normal cooking utensils and dangerous fire sources from the perspective of material properties, further improving the accuracy of heat source attribute identification and providing a reliable basis for risk assessment.

[0066] In an optional embodiment, determining the properties of the target heat source based on the flicker state, the trajectory deviation of the hot spot centroid, and the specific heat capacity of the target heat source includes: Under the condition that the first preset condition is met, the target heat source is determined to be a fire heat source. The first preset condition is that the flashing state is a high-frequency flashing state, and the trajectory deviation of the hot spot centroid is greater than or equal to a preset trajectory deviation threshold, and the specific heat capacity of the target heat source is greater than or equal to a preset specific heat capacity threshold. Under the condition that the second preset condition is met, the target heat source is determined to be a cooking heat source. The second preset condition is that the flickering state is stable, the trajectory deviation of the hot spot centroid is less than the preset trajectory deviation threshold, and the specific heat capacity of the target heat source is less than the preset specific heat capacity threshold.

[0067] Optionally, the preset trajectory deviation threshold and the preset specific heat capacity threshold can be set according to actual business needs, and are not limited here. For example, the preset trajectory deviation threshold can be 0.5 m / s, and the preset specific heat capacity threshold can be 800 J / (kg). ℃).

[0068] Optionally, determining that a target heat source is a fire requires the simultaneous fulfillment of three conditions: by analyzing the frequency domain characteristics of the infrared image through the first target model, if the output result is high-frequency flickering, it indicates that the heat source has unstable combustion characteristics, which is consistent with the dynamic characteristics of a fire; the centroid of the hot spot of the fire heat source will undergo significant displacement due to the irregularity of combustion, and its trajectory deviation exceeds a preset threshold; when the specific heat capacity of the target heat source reaches or exceeds the preset specific heat capacity threshold, it indicates that it is more likely to be a fire involving the combustion of flammable materials.

[0069] Optionally, to determine whether a target heat source is a normal cooking heat source, three conditions must be met simultaneously: First, the target model outputs a stable state, indicating that the heat source is in a continuous and stable heating state; second, the displacement of the centroid of the heat spot of the heat source such as the pot is extremely small, and its trajectory deviation is far below the preset trajectory deviation threshold, reflecting that the spatial position of the heat source is stable and there is no irregular movement like a flame; third, the specific heat capacity is less than the preset specific heat capacity threshold, indicating that the heat source is a low-heat-capacity metal material, which is consistent with the physical characteristics of cooking utensils.

[0070] The multi-dimensional feature combination judgment mechanism solves the problem that traditional systems cannot distinguish between cooking heat sources and fire conditions, resulting in a high false alarm rate. It can accurately identify fire conditions and avoid missed alarms, while eliminating interference from normal cooking and reducing false alarms. This significantly improves the accuracy of heat source attribute identification and provides a reliable basis for subsequent risk assessment and intervention.

[0071] S104. Based on the dynamic threshold and the properties of the target heat source, perform a risk assessment to obtain the risk assessment result; In an optional embodiment, the step of performing a risk assessment based on the dynamic threshold and the properties of the target heat source to obtain a risk assessment result includes: Obtain the second target model; the second target model is obtained by risk assessment prediction training of the second preset model based on historical dynamic thresholds, attributes of historical heat sources and risk assessment labels corresponding to the historical heat sources; The dynamic threshold and the attributes of the target heat source are input into the second target model to obtain the risk assessment result of the target heat source.

[0072] Optionally, the second target model is a machine learning model for predicting the risk level of heat sources, constructed based on supervised training using historical data. Historical dynamic thresholds and historical heat source attributes are used as inputs, and historical risk assessment labels are used as the supervised target to train the second pre-defined model. Through iterative optimization of the model parameters, the model learns the mapping relationship between dynamic thresholds, heat source attributes, and risk assessment results. After training, a second target model that can be directly used for real-time risk prediction is obtained.

[0073] Optionally, the real-time dynamic threshold and the attributes of the target heat source are input into the second target model. The model outputs the corresponding risk assessment result based on the mapping relationship learned during training. The risk assessment result can be a specific risk value, ranging from 0 to 1. The larger the value, the greater the risk of the fire. If the dynamic threshold is high and the heat source attribute is a fire heat source, the model will output a high-risk result. If the dynamic threshold is low and the environmental risk is small and the heat source attribute is a cooking heat source, the model will output a low-risk result.

[0074] Optionally, the formula for calculating risk assessment is: ; Where R is the risk assessment value, and a is the adjustable coefficient. The probability that the target heat source is a fire. The severity coefficient of a fire originating from a heat source. For the rate of temperature change, The first environmental compensation coefficient, This is the second environmental compensation coefficient. These are the interfering factors, namely solar radiation and airflow disturbance. The weights of the interference factors; The calculation formula is: ,in, These are the weighting coefficients. The specific heat capacity of the target heat source. The change in temperature This represents the current wind speed.

[0075] The introduction of the second objective model enables intelligent and accurate assessment of heat source risks. It comprehensively considers the impact of environmental factors and heat source attributes on risks, avoiding misjudgments caused by single indicators. The model trained based on historical data can adapt to the risk characteristics of different kitchen scenarios, improving the universality of the assessment. In addition, the output risk assessment results provide a quantitative basis for subsequent graded intervention, solving the problem of inaccurate risk assessment caused by traditional systems relying on single indicators or manually set thresholds, and improving the level of intelligence in kitchen heat source monitoring.

[0076] S105. Intervention measures are taken based on the risk assessment results.

[0077] In an optional embodiment, the intervention based on the risk assessment results includes: If the risk assessment result is less than the first risk threshold, an alarm will be pushed to the system. If the risk assessment result is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, an audible and visual alarm is triggered, and the exhaust equipment in the kitchen is activated. If the risk assessment result is greater than the second risk threshold, a telephone alarm will be triggered, and the gas appliances in the kitchen will be shut off.

[0078] Optionally, the first and second risk thresholds can be set according to actual business needs, and are not limited here. For example, if the first risk threshold is 0.3 and the second risk threshold is 0.7, when R is less than 0.3, it indicates low risk. At this time, the target heat source has a low risk and is usually in a normal cooking state. A mild warning message can be pushed to the user through an interactive module such as an APP. The content may include the current heat source temperature, environmental status, etc. No mandatory intervention is required; it is just to remind the user to pay attention. When R is greater than or equal to 0.3 and less than or equal to 0.7, it indicates medium risk. At this time, the target heat source has a medium risk, which may be due to abnormal cooking, such as the temperature of the oil pan. If the temperature is too high or the gas combustion is incomplete, but has not reached the level of a fire, a clear warning will be issued through the local sound and light device in the kitchen to attract the user's immediate attention. At the same time, the exhaust equipment will be activated, controlling the kitchen exhaust fan or range hood to accelerate the removal of fumes, high-temperature gases, or small amounts of gas, reducing environmental risks. When R is greater than 0.7, it indicates a high risk. At this time, the target heat source is at high risk, and there is a high probability that a fire has occurred. The interactive module will automatically dial the user's preset contact number to play a voice alarm message, ensuring that the user can be informed of the danger in time when away from the kitchen. The gas valve will be automatically closed to cut off the gas supply, preventing the flame from continuing to burn and delaying the spread of the fire.

[0079] For example, when stir-frying on a gas stove, the system collects and analyzes data from multiple sensors. Infrared thermal imaging data shows that the temperature of the pot bottom reaches 280℃ without flickering, indicating a stable state and confirming it as a cooking heat source, which does not exceed the dynamic threshold calculated by the system based on the current environment. Infrared thermal imaging data also identifies a hot spot with a regular shape and no flame characteristics. Millimeter-wave radar data detects that the displacement velocity of the hot spot's centroid is 0.2 m / s, lower than the preset displacement velocity threshold, and the trajectory deviation is less than the preset deviation threshold. The specific heat capacity of the pot is also less than the preset specific heat capacity threshold. Based on the above data, the heat source is ultimately identified as a cooking heat source, and a risk score of R=0.12 is calculated. Information is then pushed to the user via the app without mandatory intervention.

[0080] For example, in a high-risk fire scenario, infrared thermal imaging data detected a flame temperature of 600℃ and exhibited a high-frequency flicker of 10Hz. Analysis of the frequency domain characteristics of the infrared image using the first target model outputs a high-frequency flicker state. The high temperature of 600℃ exceeds the dynamic threshold calculated by the system based on the current environment. Visible light camera data identified an irregular hotspot shape accompanied by smoke, consistent with the visual characteristics of a fire heat source. Millimeter-wave radar data detected a displacement velocity of 1.2 m / s for the hotspot's centroid, exceeding the preset displacement velocity threshold. The high diffusion velocity resulted in a significant cumulative effect of horizontal and vertical displacement of the hotspot's centroid, with a trajectory deviation far exceeding the preset deviation threshold, reflecting the rapid spread of the flame and consistent with the fire's motion characteristics. Ultimately, this heat source was determined to be a fire heat source. At this point, Extremely high, Due to the high temperature and rapid spread, the risk score increases significantly, and the temperature change rate is extremely high. The final calculated risk score is R=0.88, which is greater than the high-risk alarm threshold. The highest level of measures are implemented, such as controlling the gas valve to shut off the fuel supply from the source and preventing the flame from continuing to burn. The interactive module triggers a telephone alarm to ensure that users can be informed of the fire immediately even if they are not in the kitchen, and take timely fire extinguishing or evacuation measures.

[0081] Optionally, the user app can overlay infrared thermal images and visible light images to mark risk areas, push operation guidelines based on fire severity levels, and support filtering fire records by timeline and exporting analysis reports. The local device can be equipped with a voice module, allowing users to terminate alarms via voice commands to avoid accidental operation, and can also include physical buttons, with an emergency stop button located on the side of the stove.

[0082] Optionally, an alarm cooling-off period can be set, where alarms of the same risk level are only pushed once every 10 minutes. When the risk level of the heat source is stable at the same level, the system will not push alarms repeatedly due to small fluctuations within that level, thereby reducing redundant alarms under the same risk state and preventing users from ignoring important information due to frequent prompts.

[0083] Optionally, when a state transition is triggered, i.e., the risk level changes from low to medium or from medium to high, a new notification is immediately sent. Regardless of whether it is in the alarm cooldown period, the system will immediately push a new alarm notification to ensure that the critical information of risk escalation is not suppressed by the cooldown period rules, so that users can grasp the worsening situation as soon as possible and take timely countermeasures.

[0084] Optionally, if a user reports an alarm via the alarm app or confirms via voice that the alarm has been handled, the same type of alarm will not be pushed again for 24 hours to avoid continued interference when the user has already intervened. At the same time, a validity period is set to ensure that the system can restore its monitoring and alarm capabilities in subsequent new risk cycles.

[0085] Without interfering with normal cooking, the system maintains the user's right to know the kitchen's status, avoids frequent and forceful interventions that could negatively impact the user experience, mitigates moderate risks through proactive intervention, prompts users to take further action to prevent escalation of risks, and controls the spread of risks through maximum intervention measures. It also ensures rapid response from users or relevant personnel to minimize fire damage. This approach, which matches different levels of intervention based on risk level, balances safety and user experience, and proactively reduces risks through interactive modules and hardware devices, rather than relying solely on manual user intervention, thus improving emergency response efficiency.

[0086] Please see Figure 2 , Figure 2 This is a structural diagram of a kitchen heat source detection system according to an embodiment of the present invention. On the other hand, this application provides a kitchen heat source detection system, which includes a data acquisition module 201, an edge computing module 202, and an execution module 203. The data acquisition module 201 is used to collect current kitchen environment information and target heat source status information; The edge computing module 202 is used to determine a dynamic threshold based on the current kitchen environment information, wherein the dynamic threshold represents the temperature threshold corresponding to the current kitchen environment information; and to determine the attributes of the target heat source based on the state information of the target heat source; and to perform a risk assessment based on the dynamic threshold and the attributes of the target heat source to obtain a risk assessment result. The execution module 203 is configured to perform push alarm processing when the risk assessment result is less than the first risk threshold; and to perform audible and visual alarm processing and control the ventilation equipment in the kitchen to start when the risk assessment result is greater than or equal to the first risk threshold and less than or equal to the second risk threshold; and to perform telephone alarm processing and control the gas equipment in the kitchen to shut down when the risk assessment result is greater than the second risk threshold.

[0087] In an optional embodiment, the edge computing module 202 includes: The dynamic threshold determination submodule is used to determine a dynamic threshold based on the current kitchen environment information; and to obtain a target radiation intensity reference value corresponding to the current solar altitude angle based on second preset information when the solar radiation intensity is greater than a preset radiation intensity threshold, wherein the second preset information represents the correspondence between the solar altitude angle and the radiation intensity reference value; and to update the solar radiation intensity to the target radiation intensity reference value and re-determine the dynamic threshold when the difference between the target radiation intensity reference value and the solar radiation intensity is greater than a preset threshold; and to obtain a target first environmental compensation coefficient corresponding to the current wind speed based on third preset information when the wind speed is greater than a preset wind speed threshold, wherein the third preset information represents the correspondence between the wind speed and the first environmental compensation coefficient; update the current first environmental compensation coefficient to the target first environmental compensation coefficient and re-determine the dynamic threshold.

[0088] The attribute determination submodule is used to determine the attributes of the target heat source based on the state information of the target heat source; The risk assessment submodule is used to obtain a second target model; the second target model is obtained by performing risk assessment prediction training on a second preset model based on historical dynamic thresholds, attributes of historical heat sources, and risk assessment labels corresponding to the historical heat sources; and is used to input the dynamic thresholds and attributes of the target heat source into the second target model to obtain the risk assessment result of the target heat source.

[0089] In an optional embodiment, the dynamic threshold determination submodule includes: An environmental change rate determination unit is used to acquire a preset benchmark threshold, a current first environmental compensation coefficient, and a second environmental compensation coefficient; and to determine the product of the temperature change rate and the humidity change rate as the environmental change rate. A radiation intensity compensation value determination unit is used to determine the sum of the solar radiation intensity and the second environmental compensation coefficient as the radiation intensity compensation value; An environmental impact factor determination unit is used to determine the ratio of the environmental change rate to the radiation intensity compensation value as the environmental impact factor. The threshold adjustment range determination unit is used to determine the product of the current first environmental compensation coefficient and the environmental impact factor as the threshold adjustment range; A threshold adjustment value determination unit is used to determine the product of the reference threshold and the threshold adjustment range as the threshold adjustment value; A dynamic threshold determination unit is used to determine the sum of the baseline threshold and the threshold adjustment value as the dynamic threshold.

[0090] In an optional embodiment, the attribute determination submodule includes: A flickering state determination unit is used to acquire a first target model; the first target model is obtained by training a first preset model to predict flickering state based on the frequency domain features of historical infrared images and the flickering state labels corresponding to the frequency domain features of the historical infrared images; and is used to input the frequency domain features of the infrared images into the first target model to obtain the flickering state corresponding to the frequency domain features of the infrared images. The trajectory deviation determination unit is used to determine the trajectory deviation of the hot spot centroid based on the displacement information of the hot spot centroid. A specific heat capacity determination unit is used to determine the specific heat capacity of the target heat source based on the absorbed heat and the temperature change. An attribute determination unit is configured to determine, under a first preset condition, that the target heat source is a fire heat source, wherein the first preset condition is that the flashing state is a high-frequency flashing state, and the trajectory deviation of the hot spot centroid is greater than or equal to a preset trajectory deviation threshold, and the specific heat capacity of the target heat source is greater than or equal to a preset specific heat capacity threshold; and to determine, under a second preset condition, that the target heat source is a cooking heat source, wherein the second preset condition is that the flashing state is a stable state, and the trajectory deviation of the hot spot centroid is less than the preset trajectory deviation threshold, and the specific heat capacity of the target heat source is less than the preset specific heat capacity threshold.

[0091] In an optional embodiment, the trajectory deviation determination unit includes: The attribute influence value determination subunit is used to obtain a preset displacement velocity threshold; and to determine the ratio of the product of the horizontal displacement and the vertical displacement to the displacement velocity threshold, which is the attribute influence value; The trajectory deviation determination subunit is used to sum the attribute influence values ​​corresponding to several time points to obtain the trajectory deviation of the hot spot centroid.

[0092] In an optional embodiment, the specific heat capacity determination unit includes: The target heat source quality determination subunit is used to acquire first preset information, which characterizes the relationship between heat sources and heat source quality; and to acquire the target heat source quality corresponding to the target heat source based on the first preset information. The specific heat capacity determination subunit is used to determine the ratio of the product of the absorbed heat and the temperature change and the mass of the target heat source, which is the specific heat capacity of the target heat source.

[0093] The kitchen heat source detection method provided in this application includes: collecting current kitchen environment information and target heat source status information; determining a dynamic threshold based on the current kitchen environment information, wherein the dynamic threshold represents a temperature threshold corresponding to the current kitchen environment information; determining the attributes of the target heat source based on the target heat source status information; performing a risk assessment based on the dynamic threshold and the target heat source attributes to obtain a risk assessment result; and performing intervention processing based on the risk assessment result. The kitchen heat source detection method provided in this application has the following beneficial effects: (1) By using multimodal sensor fusion and multidimensional technical means, accurate monitoring is achieved. Dynamic threshold calculation is used to avoid threshold rigidity caused by environmental interference. By identifying the heat source attributes, cooking heat sources and fire heat sources are distinguished. Solar radiation correction and wind speed compensation are added to further eliminate natural environmental interference, ultimately greatly reducing the false alarm rate and ensuring stable operation of the system in complex kitchen environments, reducing the number of times users shut down the system due to false alarms. (2) This case uses an edge computing module to complete dynamic threshold calculation, heat source attribute identification and risk assessment locally without relying on cloud transmission, which significantly shortens the response time; and reduces user misoperation and improves the effectiveness of emergency response through refined risk assessment and graded response mechanism. (3) The threshold and model parameters are automatically adjusted according to the environment and user habits without manual intervention. The sensor layout is accurate and the low power consumption and low cost design also supports large-scale deployment from home to industrial scenarios.

[0094] In an optional embodiment, this application provides an electronic device including a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement the kitchen heat source detection method as described in any of the above embodiments.

[0095] In an optional embodiment, this application provides a computer-readable storage medium storing at least one instruction or at least one program, which is loaded and executed by a processor to implement the kitchen heat source detection method as described in any of the above embodiments.

[0096] The above description is only an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for detecting a kitchen heat source, characterized in that, The method for detecting kitchen heat sources includes: Collect current kitchen environment information and target heat source status information; Based on the current kitchen environment information, a dynamic threshold is determined, wherein the dynamic threshold represents the temperature threshold corresponding to the current kitchen environment information; Based on the state information of the target heat source, determine the attributes of the target heat source; Based on the dynamic threshold and the properties of the target heat source, a risk assessment is performed to obtain the risk assessment result; Intervention measures will be taken based on the risk assessment results.

2. The method for detecting kitchen heat sources according to claim 1, characterized in that, The current kitchen environment information includes temperature change rate, humidity change rate, and solar radiation intensity. Determining the dynamic threshold based on the current kitchen environment information includes: Obtain the preset baseline threshold, the current first environmental compensation coefficient, and the second environmental compensation coefficient; The product of the temperature change rate and the humidity change rate is determined as the environmental change rate; The sum of the solar radiation intensity and the second environmental compensation coefficient is determined as the radiation intensity compensation value; The ratio of the environmental change rate to the radiation intensity compensation value is determined as the environmental impact factor. The product of the current first environmental compensation coefficient and the environmental impact factor is determined as the threshold adjustment range; The product of the baseline threshold and the threshold adjustment range is determined as the threshold adjustment value; The sum of the baseline threshold and the threshold adjustment value is determined as the dynamic threshold.

3. The method for detecting kitchen heat sources according to claim 1, characterized in that, The state information of the target heat source includes infrared image frequency domain features, displacement information of the hot spot centroid, absorbed heat, and temperature change. Determining the attributes of the target heat source based on its state information includes: Obtain a first target model; the first target model is obtained by training a first preset model to predict the flicker state based on the frequency domain features of historical infrared images and the flicker state labels corresponding to the frequency domain features of the historical infrared images. The infrared image frequency domain features are input into the first target model to obtain the flicker state corresponding to the infrared image frequency domain features; Based on the displacement information of the hot spot centroid, the trajectory deviation of the hot spot centroid is determined; The specific heat capacity of the target heat source is determined based on the absorbed heat and the temperature change. The properties of the target heat source are determined based on the flickering state, the trajectory deviation of the hot spot centroid, and the specific heat capacity of the target heat source.

4. The method for detecting kitchen heat sources according to claim 3, characterized in that, The hot spot centroid displacement information includes horizontal displacement and vertical displacement. Determining the trajectory deviation of the hot spot centroid based on the hot spot centroid displacement information includes: Obtain the preset displacement velocity threshold; The ratio of the product of the horizontal displacement and the vertical displacement to the displacement velocity threshold is determined as the attribute influence value; The trajectory deviation of the hot spot centroid is obtained by summing the influence values ​​of the attribute corresponding to several time points.

5. The method for detecting kitchen heat sources according to claim 3, characterized in that, Determining the specific heat capacity of the target heat source based on the absorbed heat and the temperature change includes: Obtain first preset information, which characterizes the relationship between the heat source and its mass; Based on the first preset information, the target heat source mass corresponding to the target heat source is obtained; The specific heat capacity of the target heat source is determined by the ratio of the product of the absorbed heat and the temperature change to the mass of the target heat source.

6. The method for detecting kitchen heat sources according to claim 3, characterized in that, The step of determining the properties of the target heat source based on the flickering state, the trajectory deviation of the hot spot centroid, and the specific heat capacity of the target heat source includes: Under the condition that the first preset condition is met, the target heat source is determined to be a fire heat source. The first preset condition is that the flashing state is a high-frequency flashing state, and the trajectory deviation of the hot spot centroid is greater than or equal to a preset trajectory deviation threshold, and the specific heat capacity of the target heat source is greater than or equal to a preset specific heat capacity threshold. Under the condition that the second preset condition is met, the target heat source is determined to be a cooking heat source. The second preset condition is that the flickering state is stable, the trajectory deviation of the hot spot centroid is less than the preset trajectory deviation threshold, and the specific heat capacity of the target heat source is less than the preset specific heat capacity threshold.

7. The method for detecting kitchen heat sources according to claim 1, characterized in that, The step of conducting a risk assessment based on the dynamic threshold and the properties of the target heat source to obtain a risk assessment result includes: Obtain the second target model; the second target model is obtained by risk assessment prediction training of the second preset model based on historical dynamic thresholds, attributes of historical heat sources and risk assessment labels corresponding to the historical heat sources; The dynamic threshold and the attributes of the target heat source are input into the second target model to obtain the risk assessment result of the target heat source.

8. The method for detecting kitchen heat sources according to claim 1, characterized in that, The intervention based on the risk assessment results includes: If the risk assessment result is less than the first risk threshold, an alarm will be pushed to the system. If the risk assessment result is greater than or equal to the first risk threshold and less than or equal to the second risk threshold, an audible and visual alarm is triggered, and the exhaust equipment in the kitchen is activated. If the risk assessment result is greater than the second risk threshold, a telephone alarm will be triggered, and the gas appliances in the kitchen will be shut off.

9. The method for detecting kitchen heat sources according to claim 2, characterized in that, The current kitchen environment information also includes the current wind speed and the current solar altitude angle. After determining the dynamic threshold based on the current kitchen environment information, the method further includes: When the solar radiation intensity is greater than a preset radiation intensity threshold, a target radiation intensity reference value corresponding to the current solar altitude angle is obtained based on second preset information, wherein the second preset information characterizes the correspondence between the solar altitude angle and the radiation intensity reference value. If the difference between the target radiation intensity reference value and the solar radiation intensity is greater than a preset threshold, the solar radiation intensity is updated to the target radiation intensity reference value, and the dynamic threshold is redefined. When the wind speed is greater than a preset wind speed threshold, a target first environmental compensation coefficient corresponding to the current wind speed is obtained based on third preset information, wherein the third preset information represents the correspondence between wind speed and the first environmental compensation coefficient. Update the current first environmental compensation coefficient to the target first environmental compensation coefficient, and redetermine the dynamic threshold.

10. A kitchen heat source detection system, characterized in that, The kitchen heat source detection system includes a data acquisition module, an edge computing module, and an execution module; The data acquisition module is used to collect current kitchen environment information and target heat source status information; The edge computing module is used to determine a dynamic threshold based on the current kitchen environment information; And for determining the attributes of the target heat source based on the state information of the target heat source; And to perform risk assessment based on the dynamic threshold and the properties of the target heat source, and obtain risk assessment results; The execution module is used to perform intervention based on the risk assessment results.

Images