Air conditioner

Abstract

The invention discloses an air conditioner, and belongs to the technical field of air conditioners. The air conditioner comprises an indoor unit and an outdoor unit, wherein an air outlet is formed in a shell; the thermal imaging device is arranged on the air outlet and used for collecting a thermal imaging image of the infrared detection area; the air guide plate is arranged at the air outlet, and the air guide plate comprises a transverse air guide plate and a longitudinal air guide plate; and the controller is configured to obtain a thermal imaging image, perform hot spot clustering on the thermal imaging image, identify a head area, a waist area and distance information, control the swing angles of the transverse air deflector and the longitudinal air deflector based on the head area, the waist area and the distance information, and switch an air supply mode of the air conditioner into a blowing mode or a avoiding mode. According to the air conditioner, accurate positioning and tracking can be achieved, and intelligent air supply is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of air conditioners, in particular to an air conditioner. BACKGROUND

[0002] In the existing air conditioner air supply control technology, intelligent air supply is often realized by positioning the position of the human body to improve the comfort experience and energy efficiency performance of the user. However, the existing technology still has obvious deficiencies in the accuracy of human body positioning, target resolution capability and privacy protection, which limits the further optimization of system performance.

[0003] At present, a common scheme is to realize human body positioning based on an infrared sensor. Although this scheme has advantages such as non-contact and no dependence on light, it usually only uses a single infrared detection device, ignoring the significant influence of human body posture information (such as standing and sitting posture) on the projected area of thermal radiation. Experimental data shows that the thermal imaging area of the human body in standing and sitting states can differ by 40%, thereby causing the thermal center calculation to deviate, affecting the positioning accuracy. In addition, under the condition that the distance between the human body and the sensor is more than 3 meters, the positioning error of the existing infrared scheme often exceeds ±30 cm, which is difficult to meet the fine air supply control demand.

[0004] On the other hand, millimeter wave radar is widely used in indoor target detection due to its low privacy sensitivity. However, the existing technology mostly uses 24 GHz frequency band millimeter wave radar, which has a horizontal beam width of 80° and a low angle resolution, making it difficult to effectively distinguish adjacent individuals, especially in a space where multiple people are present, which can easily cause target confusion, affecting the accuracy and stability of positioning and tracking.

[0005] In addition, there are existing schemes that attempt to introduce video analysis based on RGB cameras to identify human body parts, thereby further optimizing the air supply strategy. However, this method involves image collection and identification of facial and body features, which poses a serious risk of privacy leakage, especially in environments such as homes and offices where privacy requirements are high, making it difficult for users to widely accept. SUMMARY

[0006] In a first aspect, the present application provides an air conditioner, comprising: an indoor unit, a housing of which is provided with an air outlet; a thermal imaging device, disposed on the air outlet, for collecting thermal imaging images of an infrared detection area; a deflector, disposed at the air outlet, the deflector comprising a transverse deflector and a longitudinal deflector; a controller, disposed on the indoor unit, the controller being electrically connected to the thermal imaging device, the controller being configured to: acquire a thermal image of the infrared detection area by the thermal imaging device, perform thermal spot clustering on the thermal image, identify a target area and a background area in the thermal image, and identify a head area and a waist area, the thermal image containing temperature values of respective pixel points; calculate a first distance of the target area based on a maximum thermal gradient difference of the target area and the background area, determine distance information of the target area based on the first distance, and correct the distance information to obtain corrected distance information; match posture information of the target object based on the head area and the waist area, control swing angles of the transverse air deflector and the longitudinal air deflector according to the posture information and the corrected distance information, and switch a blowing mode of the air conditioner to a human blowing mode or a human avoiding mode.

[0007] Compared with a traditional infrared human body detection method, the use of thermal imaging and thermal spot clustering can more finely identify the positions and postures of different parts of a human body, and improve the accuracy and response speed of switching of the blowing mode. The distance correction method based on a thermal gradient difference effectively reduces the distance measurement error caused by factors such as light and changes in ambient temperature, and improves the stability of the system in a complex indoor environment. At the same time, automatic switching of the human blowing mode and the human avoiding mode can meet the comfort needs of different users and reduce unnecessary energy consumption.

[0008] In some embodiments, the air conditioner further includes: a radar sensor arranged on the air outlet and configured to detect a second distance of the target area; The determination of the distance information of the target area based on the first distance and the correction includes: calculate a thermal imaging confidence based on a temperature difference between the target area and the background area, determine a radar confidence based on a multipath interference index, perform weighted calculation on the first distance and the second distance based on the thermal imaging confidence and the radar confidence, and perform normalization to obtain the distance information.

[0009] Based on the above configuration, the controller quantifies the reliability of the thermal image by calculating the thermal imaging confidence, quantifies the reliability of radar distance measurement according to the multipath interference index of the radar signal, performs weighted fusion according to the corresponding confidence weights, and processes the distance information through normalization to realize high-precision fusion of multi-source distance measurement data.

[0010] In some embodiments, the determination of the distance information of the target area based on the first distance and the correction further includes: construct a feature factor based on the distance information, a horizontal deflection angle, and a vertical deflection angle of the thermal imaging device, calculate the feature factor based on an elastic network regression model, and obtain the corrected distance information.

[0011] In some embodiments, the controller is further configured to: identify head centroid coordinates of the head region and identify waist centroid coordinates of the waist region, calculate displacement amounts of the head centroid coordinates, the waist centroid coordinates in consecutive multiple thermal imaging images, and predict the head centroid coordinates, the waist centroid coordinates at the next time based on the displacement amounts using a linear extrapolation model; control swing angles of the transverse air deflector, the longitudinal air deflector at the next time based on the predicted head centroid coordinates, the waist centroid coordinates.

[0012] Based on the above configuration, the controller can calculate the centroid coordinates of the head and the waist at the next time according to the historical displacement information by using the linear extrapolation model, so as to realize short-time prediction of the human body position. The predicted position data will be used to calculate the swing angles of the transverse air deflector and the longitudinal air deflector in advance, so that the air supply direction can be adjusted in time while the human body moves.

[0013] In some embodiments, the thermal spot clustering of the thermal imaging image, the identification of the target region and the background region in the thermal imaging image, and the identification of the head region and the waist region further comprise: extracting pixel points with temperature values greater than a human body metabolism temperature threshold in the thermal imaging image as high-temperature points, clustering the high-temperature points based on a clustering algorithm DBSCAN, and identifying a plurality of thermal spot regions as target regions; determining a minimum rectangular region containing the thermal spot region as a target circumscribed region, expanding the width and height of the target circumscribed region by a specified multiple outwardly, and excluding pixel points with temperature values greater than a noise temperature threshold in the expanded region to obtain a background region; calculating an elliptical fitting degree of the thermal spot region, and if the elliptical fitting degree exceeds a set elliptical threshold and the temperature standard deviation of the thermal spot region is lower than a temperature difference threshold, marking the corresponding thermal spot region as a head region; calculating a vertical temperature gradient of the thermal spot region, and if there is a continuous pixel column satisfying the vertical temperature greater than a gradient threshold, marking the corresponding thermal spot region as a waist region.

[0014] Based on the above configuration, combined with the use of temperature threshold screening, density clustering and geometric feature analysis, the interference of environmental heat sources on human body recognition can be effectively reduced. The DBSCAN clustering algorithm does not need to preset the number of clusters, and can automatically identify heat source regions of any shape, which is suitable for application scenarios with variable human body postures. By introducing the double conditions of elliptical fitting degree and temperature standard deviation for screening, the head position can be accurately located; and the use of vertical temperature gradient to determine the waist improves the reliability of human body posture recognition.

[0015] In some embodiments, the calculating the first distance of the target region based on the maximum thermal gradient difference between the target region and the background region further comprises: calculating a difference between a highest temperature value in the target region and an average temperature value of the background region to obtain the maximum thermal gradient difference; performing a power operation on the maximum thermal gradient difference with the thermal radiation attenuation coefficient as an index, taking a ratio of a product of the power operation result and a focal length of the thermal imaging device and a product of a calibration constant of the thermal imaging device and an alignment deviation angle as the first distance, wherein the alignment deviation angle is a pixel coordinate deviation angle of a center of mass coordinate of the target region and an optical axis center.

[0016] Based on the above configuration, the present embodiment combines the maximum thermal gradient difference with the thermal radiation attenuation model, so that the distance measurement process not only depends on the temperature difference, but also considers the physical law of infrared radiation attenuation with distance, thereby maintaining high distance measurement accuracy under different environmental temperature and humidity conditions.

[0017] In some embodiments, the calculating the thermal imaging confidence based on the temperature difference between the target region and the background region further comprises: taking the temperature difference between the target region and the background region as a signal amplitude, taking a standard deviation of the temperature measurement noise as a noise amplitude, and calculating a temperature signal-to-noise ratio using an amplitude signal-to-noise ratio operation model based on the signal amplitude and the noise amplitude; taking a negative value of the temperature signal-to-noise ratio after being multiplied by a certain number, taking an exponential function with a natural constant as an index, and taking a difference between 1 and the exponential function as the thermal imaging confidence.

[0018] Compared with the conventional method of determining effectiveness based only on a temperature difference threshold, the present method can more sensitively reflect the influence of environmental noise on imaging effect, thereby more accurately determining the confidence level of the thermal imaging image in a complex indoor environment (such as air flow disturbance of an air conditioner, direct sunlight, etc.). By introducing the confidence calculation result, the weight can be adaptively adjusted in subsequent multi-sensor data fusion, thereby improving the accuracy and stability of overall distance measurement and human posture recognition.

[0019] In some embodiments, the controller is further configured to: monitoring a vertical temperature gradient of the waist region of the continuous frames, and if the vertical temperature gradient increases, identifying the posture information as switching from a sitting posture to a standing posture, and switching the air supply mode of the air conditioner to a human-avoiding mode; if the vertical temperature gradient decreases, identifying the posture information as switching from a standing posture to a sitting posture, and switching the air supply mode of the air conditioner to a human-approaching mode.

[0020] Based on the above configuration, the dynamic change of the vertical temperature gradient of the waist region is utilized to determine the posture switching of the human body. The controller identifies the waist region in the thermal imaging image and calculates the temperature gradient value of the region along the vertical direction in continuous multiple frames of images.

[0021] In some embodiments, the controller is further configured to: determine a horizontal orientation angle of the human body based on the ratio of the horizontal coordinate of the centroid coordinate of the head region to the distance information, and an inverse tangent function of the ratio; determine a vertical orientation angle of the human body based on the product of the difference between the vertical coordinate of the centroid coordinate of the head region and the vertical coordinate of the centroid coordinate of the waist region and a set proportionality coefficient; control the swing angle of the transverse air deflector and the longitudinal air deflector according to the horizontal orientation angle and the vertical orientation angle of the human body.

[0022] In a second aspect, the present application provides an air conditioner, comprising: an indoor unit, a housing of which is provided with an air outlet; a thermal imaging device, arranged on the air outlet, for acquiring a thermal imaging image of an infrared detection region; a radar sensor, arranged on the air outlet, for detecting a second distance of a target region; an air deflector, arranged at the air outlet, the air deflector comprising a transverse air deflector and a longitudinal air deflector; a controller, arranged on the indoor unit, the controller being electrically connected to the thermal imaging device, the controller being configured to control the action of the air deflector by implementing an air conditioner control method, the control method comprising: a target identification step, acquiring a thermal imaging image of the infrared detection region, performing thermal spot clustering on the thermal imaging image, identifying a target region and a background region in the thermal imaging image, and identifying a head region and a waist region; a distance acquisition step, calculating a first distance of the target region based on the maximum thermal gradient difference of the target region and the background region, and performing weighted calculation on the first distance based on the second distance to obtain distance information; and a blowing control step, based on the posture information of the target object matched by the head region and the waist region, controlling the swing angle of the transverse air deflector and the longitudinal air deflector according to the posture information and the corrected distance information, and switching the blowing mode of the air conditioner to a human blowing mode or a human avoiding mode.

[0023] Based on the above steps, through the fusion of thermal imaging and millimeter wave radar, all-weather and high-precision human body detection and ranging is realized, even in the case of insufficient light, no obvious temperature difference or partial occlusion, it can also maintain a high recognition rate. The posture information is obtained by using the relative position relationship between the head and the waist, which can quickly switch the air supply mode when the user's action changes (such as sitting down, standing up), improve the comfort and reduce energy waste. The air deflector control combined with human azimuth angle calculation and closed-loop motor adjustment significantly improves the accuracy and response speed of the air supply direction, enabling the air conditioner to have the ability of intelligent tracking or active wind avoidance. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 A mechanical structure schematic diagram of an air conditioner is provided for the embodiments of the present application. Figure 2 A structure schematic diagram of an air outlet of an air conditioner is provided for the embodiments of the present application Figure 3 A circuit system architecture diagram of an air conditioner is provided for the embodiments of the present application. Figure 4 A logic schematic diagram of an air conditioner is provided for the embodiments of the present application. Figure 5 A temperature measurement principle diagram of a thermal imaging device is provided for the embodiments of the present application. Figure 6 A calculation logic schematic diagram of an alignment deviation angle is provided for the embodiments of the present application. Figure 7 A calculation logic schematic diagram of distance information is provided for the embodiments of the present application. Figure 8 A correction logic schematic diagram of distance information is provided for the embodiments of the present application. Figure 9 A logic schematic diagram of a controller is provided for the embodiments of the present application. Figure 10 A flowchart of a control method is provided for the embodiments of the present application. Figure 11 A hardware structure schematic diagram of a controller is provided for the embodiments of the present application.

[0025] In the above figures: 10, air conditioner; 101, indoor unit; 102, outdoor unit; 103, pipeline; 104, air outlet; 105, thermal imaging device; 106, radar sensor; 107, controller; 108, horizontal air deflector; 109, vertical air deflector; 110, processor; 111, memory; 112, communication interface; 113, bus. DETAILED DESCRIPTION

[0026] In order to make the purposes and embodiments of the present application clearer, the following will clearly and completely describe the exemplary embodiments of the present application with reference to the accompanying drawings of the exemplary embodiments of the present application. Obviously, the described exemplary embodiments are only a part of the embodiments of the present application, but not all the embodiments.

[0027] Unless otherwise defined, technical terms or scientific terms used in the present application shall have the ordinary meaning understood by a person with ordinary skill in the art to which the present application pertains. The terms "a", "an", "one", "this", and similar terms used in the present application do not represent quantity limitation, but can represent singular or plural. The terms "include", "contain", "have", and any variation thereof used in the present application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device including a series of steps or modules (units) is not limited to the listed steps or units, but can further include steps or units not listed, or can further include other steps or units inherent to the process, method, product, or device. The terms "connect", "connected", "couple", and similar terms used in the present application are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The term "multiple" in the present application refers to two or more. The term "and / or" describes the association relationship of the associated objects, which means that there can be three relationships, for example, "A and / or B" can represent the following three cases: A exists alone, A and B exist together, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects. The terms "first", "second", "third", and the like in the present application are only to distinguish similar objects, and do not represent a specific order for the objects.

[0028] Since the thermal imaging image does not have obvious color information relative to the color image, it will not infringe on the privacy of the user in daily life, and the thermal imaging image itself is generated based on different temperatures of different objects in the infrared detection area, and based on the thermal imaging image, whether there is a target object in the infrared detection area can be accurately identified.

[0029] In this way, the temperature measurement accuracy of the thermal imaging device is improved without involving the privacy of the user, and then the operation of the air conditioning system is intelligently controlled based on the accurate temperature to meet the user demand.

[0030] In the embodiments of the present application, the thermal imaging device is an electronic device for target detection using infrared radiation, for example, a thermoelectric pile, a thermal imager, etc. In some embodiments of the present application, a thermal imager with strong resolution and good imaging effect is used. The infrared radiation can be referred to as infrared light, infrared ray, which is an electromagnetic wave with a wavelength of about 0.75 microns to 1000 microns.

[0031] Figure 1A mechanical structure schematic diagram of an air conditioner is provided for the embodiments of the present application. As shown in Figure 1 The air conditioner 10 can include: An indoor unit 101, taken as an indoor hanging machine (shown in Figure 1 The indoor hanging machine is usually installed on an indoor wall surface or the like, and an air outlet is provided on the shell thereof. An outdoor unit 102 is usually arranged outdoors and is used for indoor environment heat exchange.

[0032] The indoor unit 101 and the outdoor unit 102 are connected by a pipeline 103 for refrigerant flow.

[0033] The indoor unit 101 includes a shell. The shell is used to constitute the outer contour of the indoor unit 101 and accommodate the internal components of the indoor unit 101.

[0034] The shell has an air inlet. The air inlet is used for indoor air to enter the shell.

[0035] Figure 2 A structure schematic diagram of an air outlet of the air conditioner is shown in Figure 2 The shell has an air outlet 104. The air outlet 104 is used for the shell to discharge air. Indoor air enters the shell through the air inlet and is blown out from the air outlet 104.

[0036] The air outlet 104 can include a guide vane, which includes a transverse guide vane 108 and a longitudinal guide vane 109.

[0037] The indoor unit 101 includes an indoor fan. The indoor fan is installed in the shell, and the indoor fan rotates to make indoor air enter the indoor shell, and the indoor air flows out of the indoor shell after heat exchange with the indoor heat exchanger.

[0038] Generally, the air conditioner 10 is also configured with a remote controller, which has a function of communicating with the air conditioner 10 by using infrared rays or other communication methods, for example. The remote controller is used to realize the interaction between the user and the air conditioner 10. The user can perform operations such as air conditioner 10 switching, temperature setting, wind direction setting, and air volume setting through the display device and buttons on the remote controller.

[0039] Figure 3 A circuit system architecture diagram of the air conditioner is shown in Figure 2 to Figure 3 The air conditioner 10 further includes a thermal imaging device 105 and a controller 107, wherein the thermal imaging device 105 is arranged on the air outlet 104 and is used to collect a thermal imaging image of an infrared detection area, and a guide vane is arranged at the air outlet 104 of the indoor unit 101. The controller 107 is arranged on the indoor unit 101, and the controller 107 is electrically connected to the thermal imaging device 105.

[0040] Figure 4This is a schematic diagram of the controller's control logic, for reference. Figure 4 As shown, controller 107 is configured as follows: The thermal imaging device 105 acquires thermal imaging images of the infrared detection area, performs hot spot clustering on the thermal imaging images, identifies the target area and background area in the thermal imaging images, and identifies the head area and waist area. The thermal imaging images contain the temperature values ​​of each pixel. The first distance of the target region is calculated based on the maximum thermal gradient difference between the target region and the background region. The distance information of the target region is determined and corrected based on the first distance to obtain the corrected distance information. Based on the posture information of the target object matched in the head and waist regions, the swing angle of the horizontal and vertical air guides is controlled according to the posture information and the corrected distance information, and the air supply mode of the air conditioner 10 is switched to the blowing mode or the avoiding mode.

[0041] In some embodiments, the posture information of the target object is determined as sitting or standing based on the vertical distance ratio of the head area and the waist area. If it is sitting, the air supply mode of the air conditioner 10 is controlled to be the person avoidance mode; if it is standing, the air supply mode of the air conditioner 10 is controlled to be the person blowing mode.

[0042] In another embodiment, the corrected distance information can also be used to control the air supply speed of the air conditioner 10, with the air speed increasing as the distance increases, or the air supply speed of the air conditioner 10 can be reduced when the posture information of the target object is determined to be a sitting posture.

[0043] In another embodiment, when multiple human bodies are detected, the target object closest to the air conditioner 10 is selected as the primary target for air supply mode control.

[0044] Based on the above configuration, by integrating a thermal imaging device 105 on the indoor unit 101, the air conditioner 10 can acquire thermal imaging images of the infrared detection area in real time, with each pixel in the image corresponding to the actual temperature value. The controller 107 uses a hot spot clustering algorithm to divide the image into a target area and a background area, and further distinguishes the head area and the waist area, thereby realizing the identification of human body parts. The first distance is calculated by the maximum thermal gradient difference between the target area and the background area, and the accurate distance information is obtained by combining the geometric correction method. The controller 107 then judges the posture characteristics of the target object based on the identified head and waist position relationship, and dynamically adjusts the swing angle of the horizontal and vertical air guides by combining the corrected distance data, so as to realize the intelligent switching of the air supply mode between the blowing mode and the avoiding mode.

[0045] Compared to traditional infrared human detection methods, thermal imaging and hot spot clustering can more accurately identify the position and posture of different parts of the human body, improving the accuracy and response speed of airflow mode switching. The distance correction method based on thermal gradient difference effectively reduces ranging errors caused by factors such as lighting and ambient temperature changes, enhancing the system's stability in complex indoor environments. Simultaneously, automatic switching between blowing and avoiding human modes can meet the comfort needs of different users and reduce unnecessary energy consumption.

[0046] In the above embodiments, reference is made to Figure 5 The diagram shows the temperature measurement principle of a thermal imaging device. (Reference) Figure 5 As shown, the infrared rays emitted by the target object are focused onto the detector by the lens, and the detector converts the received thermal energy into an electrical signal output. Data transmitted between the detector and the main control board includes raw data (14-bit), the baffle temperature, and the detector temperature. Thermal imaging can be achieved using the raw data, and temperature measurement can be achieved using the raw data, baffle temperature, and detector temperature. Here, T1 represents the detector's displayed temperature; T2 represents the baffle temperature, which can be understood as the ambient temperature. By placing the thermal imaging device in a constant-temperature chamber to ensure a constant temperature, the ambient temperature measured by the thermal imaging device can be used instead in actual use.

[0047] In another embodiment, the thermal imaging device 105 can be replaced with an infrared temperature array sensor with mid-to-long-range resolution to reduce costs. The infrared temperature array sensor contains an array of multiple infrared detection units (pixels), each unit capable of measuring the infrared radiation intensity in its corresponding field of view. The infrared signal is converted into a temperature value for the corresponding pixel using the relationship between radiation intensity and temperature (Planck's law and Stefan-Boltzmann law).

[0048] In another embodiment, the air outlet 104 further includes an actuator electrically connected to the controller 107 and drivingly connected to the horizontal air guide plate 108 and the vertical air guide plate 109. Optionally, the actuator can be a stepper motor or a servo motor. Using a servo motor can improve the adjustment accuracy.

[0049] Based on the above configuration, after calculating the target angle, the controller 107 converts the angle into motor rotation steps or PWM pulse width signal, and performs angle calibration through a position sensor (such as a Hall sensor or photoelectric encoder) or a timing control method to achieve precise positioning. In the blowing mode, the controller 107 directs the air guide plate towards the head or waist area of ​​the human body; in the avoiding mode, it deviates the air guide plate from the direction of the human body at a certain angle to deliver air, thereby achieving personalized and comfortable air delivery.

[0050] By introducing a closed-loop control method that combines motor drive and position feedback, the air guide plate can achieve precise positioning and rapid response, improving the flexibility and reliability of airflow direction adjustment.

[0051] In some embodiments, calculating a first distance to the target region based on the maximum thermal gradient difference between the target region and the background region further includes: The maximum thermal gradient difference is obtained by calculating the difference between the highest temperature value in the target area and the average temperature value in the background area. It is used to reflect the difference in thermal radiation between heat sources such as the human body and the background; With thermal radiation attenuation coefficient The exponent is raised to the power of the maximum thermal gradient difference. The product of the power and the focal length of the thermal imaging device is then multiplied by the calibration constant of the thermal imaging device. and alignment deviation angle The ratio of their products is taken as the first distance, where the alignment deviation angle is... The centroid coordinates of the target region Center of optical axis Pixel coordinate deviation angle, optical axis center The pixel coordinate deviation angle is an inherent parameter of the thermal imaging module, and the target area can be the head area or the waist area.

[0052] In the above embodiments, the first distance is Among them, the thermal radiation attenuation coefficient It is negatively correlated with ambient humidity, with a default setting of 0.92 and a value range of 0.6 to 1.2.

[0053] Based on the above configuration, this implementation method combines the maximum thermal gradient difference with the thermal radiation attenuation model, so that the ranging process not only depends on the temperature difference, but also takes into account the physical law of infrared radiation attenuation with distance. Therefore, it still maintains high ranging accuracy under different ambient temperature and humidity conditions.

[0054] refer to Figure 6 As shown in the above embodiment, the alignment deviation angle The initial values ​​are obtained based on the following calculation model: .

[0055] In some embodiments, reference Figure 3 As shown, the air conditioner 10 also includes: Radar sensor 106, mounted on air outlet 104, is used to detect the second range of the target area. Optionally, the radar sensor 106 is disposed on one side of the thermal imaging device 105, for example as follows: Figure 2 The area within the dashed frame is encapsulated within the panel of the air outlet 104; refer to Figure 7 As shown, the distance information of the target area is determined and corrected based on the first distance, including: The thermal imaging confidence level Qir is calculated based on the temperature difference between the target area and the background area, with a value ranging from 0 to 1. The radar confidence level is then determined based on the multipath interference index MPI. Based on thermal imaging confidence and radar confidence, the first distance Second distance After weighted calculation, normalization is performed to obtain the distance information Z.

[0056] In some embodiments, the sum of the thermal imaging confidence and the radar confidence is configured to be greater than or equal to a fusion threshold, which is configured to be 1.5.

[0057] Specifically, the distance information Z is obtained based on the following calculation model: .

[0058] Among them, radar confidence Based on the following calculation model: The multipath interference index (MPI) is calculated based on the radar echo attenuation rate and ranges from 0 to 1.

[0059] In the above embodiments, when the MPI exceeds 0.3, it indicates that there is significant multipath reflection, the radar confidence decreases, and thermal imaging images are used preferentially.

[0060] In the above embodiments, when the ambient temperature fluctuates by more than 3°C, the exponential decay mechanism of the thermal imaging confidence score Qir is triggered. The decay process is calculated based on the following calculation model: t represents the duration of ambient temperature fluctuations exceeding 3°C, expressed in minutes.

[0061] Based on the above configuration, the controller 107 quantifies the reliability of the thermal imaging image by calculating the thermal imaging confidence metric, quantifies the reliability of the radar ranging according to the multipath interference index of the radar signal, performs weighted fusion according to the corresponding confidence weights, and processes the distance information through normalization to achieve high-precision fusion of multi-source ranging data.

[0062] In the above embodiments, the introduction of radar sensor 106 provides additional ranging information in low-light, obstructed, or weak temperature gradient environments, improving the robustness of the air conditioner under complex indoor conditions. This is achieved by calculating the thermal imaging confidence level (Qir) and the radar confidence level. This enables dynamic weight allocation between the two ranging methods, thereby automatically selecting the more reliable data source in different environments.

[0063] This application introduces the azimuth angle of radar sensor 106. Alignment deviation angle The initial value is corrected, and the azimuth angle is calculated by the beamforming algorithm, antenna array, and digital signal processing inside the radar sensor 106. The azimuth angle can be obtained by parsing the radar data. .

[0064] Specifically, based on the azimuth angle of radar sensor 106 The initial value of the alignment deviation angle θ is corrected, specifically based on the following model: initial values ​​and azimuth angle Perform weighted fusion: .

[0065] Optionally, considering that thermal imaging detection has higher angular resolution at close range, while radar sensor 106 has stronger anti-jamming capabilities at long range, therefore, The settings are 0.7 and 0.3. "Close distance" and "far distance" are relative to a distance of 3 meters. A distance greater than 3 meters is considered close distance; otherwise, it is considered far distance.

[0066] By introducing alignment deviation angle compensation, parallax errors caused by the target not being centered on the optical axis can be effectively reduced, especially when the air conditioner is installed at a high position or under overhead inspection. Compared with the simple temperature difference-distance linear fitting method, this method has higher robustness and accuracy in distance measurement and can maintain good consistency under different installation environments.

[0067] In the above embodiments, the focal length and calibration constant can be obtained through factory calibration or on-site adaptive calibration, and the calculation method of the alignment deviation angle can be replaced by a pixel-angle mapping relationship based on the perspective geometry model to adapt to different optical lens characteristics.

[0068] In some embodiments, calculating the thermal imaging confidence level based on the temperature difference between the target area and the background area further includes: The temperature difference between the target area and the background area As the signal amplitude, the standard deviation of the temperature measurement noise As the noise amplitude, the signal amplitude and noise amplitude are calculated using an amplitude signal-to-noise ratio (SNR) calculation model to obtain the temperature signal-to-noise ratio (SNR). ; The negative value of the temperature signal-to-noise ratio (SNR) is reduced by a factor of two. An exponential function with the natural constant as the base is taken, and the difference between 1 and the exponential function is taken as the confidence level of thermal imaging.

[0069] In practical applications, the confidence level of thermal imaging is obtained based on the following calculation model: Based on the above formula, when the SNR exceeds 20dB, the thermal imaging confidence level is approximately 1. When the SNR is less than 10dB, the radar confidence level is set to be greater than 0.8. Then, the value is calculated according to the radar confidence level calculation model. If the calculated radar confidence level is less than 0.5, the radar confidence level is set to 0.5.

[0070] Based on the above configuration, the controller 107 uses the temperature difference between the target area and the background area as the signal amplitude, which reflects the significance of the thermal difference between the target and the background in the infrared image; then, it obtains the standard deviation of the temperature measurement noise. The noise amplitude is used as the basis for calculating the temperature signal-to-noise ratio using an amplitude signal-to-noise ratio calculation model, thus determining the strength of the target signal relative to the background noise under the current imaging conditions.

[0071] Subsequently, the thermal imaging confidence score Qir was calculated based on the above model. This formula ensures that the confidence score is close to 1 when the temperature difference is large and the noise is small, while the confidence score is close to 0 when the temperature difference is small or the noise is large, thus realizing a quantitative assessment of the reliability of thermal imaging.

[0072] Compared to traditional methods that rely solely on temperature difference thresholds to determine validity, this approach is more sensitive to the impact of environmental noise on imaging results, thus enabling more accurate assessment of the reliability of thermal imaging images in complex indoor environments (such as air conditioning turbulence and direct sunlight). By incorporating this confidence score calculation, weights can be adaptively adjusted in subsequent multi-sensor data fusion, improving the overall accuracy and stability of ranging and human pose recognition.

[0073] The above embodiments ensure the stability and consistency of the final distance information Z through weighted fusion and normalization processing, effectively reducing the error caused by a single sensor malfunction, providing more accurate human position input for the air conditioner's air guide plate control, making the switching between the blowing mode and the avoidance mode more precise, and further improving comfort and energy-saving effects.

[0074] In the above embodiments, the radar sensor 106 may be a millimeter-wave radar module in the 24GHz, 60GHz or 77GHz frequency band to adapt to different ranging range and accuracy requirements.

[0075] In some embodiments, reference Figure 8 As shown, determining and correcting the distance information of the target area based on the first distance also includes: Based on distance information Z and the horizontal deflection angle of thermal imaging device 105 Vertical deflection angle Constructing feature factors The feature factors are calculated based on the elastic network regression model to obtain the corrected distance information.

[0076] In the above embodiments, the resolution of the thermal imaging device is assumed to be W120 * G90, the horizontal field of view (HFOV) is 90°, and the vertical field of view (VFOV) is 67.5°. The calculation methods for the horizontal and vertical deflection angles are as follows: The value range is [-45°, 45°].

[0077] The value range is [-33.75°, 33.75°].

[0078] In the above embodiment, the corrected distance information is: .

[0079] Among them, coefficient The measurement data, including feature factors and target labels, is obtained through a linear regression machine learning algorithm based on a large amount of measurement data. The target labels are the corrected distances obtained from laboratory measurement data. During the linear regression fitting process based on the linear regression machine learning algorithm, the regularization parameter is set to 0.1~0.5. When the mixing parameter in the elastic network is 1, the model degenerates into LASSO; when the mixing parameter is 0, the model degenerates into ridge regression. The mixing parameter in this application is set to 0.5.

[0080] The elastic network regression model combines the advantages of L1 and L2 regularization, which can maintain the sparsity of the model to improve computational efficiency and prevent overfitting, thus maintaining high generalization ability under multiple environmental conditions.

[0081] Based on the above configuration, a secondary error correction is added based on feature factors and an elastic network regression model. The feature factors include distance information Z and horizontal deflection angle. Vertical deflection angle and the square of the distance information Z 2 Horizontal deflection angle With vertical deflection angle The product, Z, is based on the square of the distance information. 2 By eliminating the inherent nonlinear bias of the sensor and utilizing the cross-coupling effect, deviations caused by factors such as equipment installation angle, distance measurement blind zone, and optical distortion can be compensated, further improving the distance measurement accuracy and thus providing more reliable input data for the control of the air guide plate.

[0082] In another embodiment, the elastic network regression model can be replaced with support vector regression (SVR), random forest regression, or deep neural networks to cope with nonlinear error characteristics. The construction of feature factors can include auxiliary parameters such as ambient temperature and humidity to enhance the model's adaptability to changing environments.

[0083] In some embodiments, thermal imaging images are subjected to hotspot clustering to identify target regions and background regions in the thermal imaging images, and head and waist regions are identified, further including: Pixels with temperatures exceeding the human metabolic temperature threshold in thermal imaging images are extracted as high-temperature points, such as, but not limited to, those around 34℃~36℃. These high-temperature points are then clustered using the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm. Each high-temperature point is represented by a hot spot, and multiple hot spot regions are identified as target areas. Each cluster contains at least eight high-temperature points to eliminate noise points. The neighborhood radius of each point is 15 cm to determine the density reachability of the points, effectively identifying areas with abnormal temperatures. Determine the smallest rectangular region containing the hot spot area as the target bounding region, expand the width and height of the target bounding region outward by a specified multiple, and exclude pixels in the expanded region whose temperature value is greater than the noise temperature threshold to obtain the background region; Calculate the ellipse fit of the hot spot region. If the ellipse fit exceeds a set ellipse threshold and the temperature standard deviation of the hot plate region is lower than the temperature difference threshold, then mark the corresponding hot spot region as the head region. Calculate the vertical temperature gradient of the hot spot region If there exists a continuous column of pixels whose vertical temperature is greater than the gradient threshold, such as 0.5℃ / cm, then the corresponding hot spot area is marked as the waist region.

[0084] In the above embodiments, considering that the head is approximately spherical and has a uniform temperature distribution, the head region satisfies the following constraints: ,and Based on this constraint, interference from environmental heat sources (such as lamps, water heaters, etc.) can be effectively distinguished.

[0085] Calculate the elliptic fit E of the hot spot region, where the ratio of the major axis a to the minor axis b ∈ [1.0, 1.5]; if E ≥ 0.85 and the temperature standard deviation is within a certain range... If the temperature is ≤ 0.8℃, it is marked as a hot zone on the head.

[0086] In another embodiment, waist positioning recognition further includes: Search for abrupt temperature gradient changes along the vertical y-axis, with the gradient symmetry difference between the left and right sides being less than 20%.

[0087] Based on the above configuration, by combining temperature threshold filtering, density clustering, and geometric feature analysis, the interference of environmental heat sources on human body recognition can be effectively reduced. The DBSCAN clustering algorithm does not require a preset number of clusters and can automatically identify heat source regions of arbitrary shapes, making it suitable for application scenarios with varying human postures. By introducing dual conditions of ellipse fit and temperature standard deviation for filtering, the head position can be located relatively accurately; while using the vertical temperature gradient to determine the waist improves the reliability of human posture recognition.

[0088] In some embodiments, reference Figure 9 As shown, controller 107 is also configured to: The vertical temperature gradient in the waist region of consecutive frames is monitored to determine whether the vertical temperature gradient increases or decreases. If the vertical temperature gradient increases, the posture information is identified as switching from sitting to standing, and the air supply mode of the air conditioner 10 is switched to the blowing mode for people. If the vertical temperature gradient decreases, the posture information is identified as switching from standing to sitting, and the air supply mode of the air conditioner 10 is switched to the avoiding mode for people.

[0089] Based on the above configuration, the dynamic changes in the vertical temperature gradient of the waist region are used to determine the change in human posture. The controller 107 identifies the waist region in the thermal imaging image and calculates the temperature gradient value of the region along the vertical direction in multiple consecutive frames.

[0090] When the human body changes from a sitting to a standing position, the temperature distribution in the lumbar region will change significantly due to the following factors: (1) Changes in body surface coverage: When sitting, the waist is in contact with the chair, local airflow decreases, and heat is easily accumulated; when standing, the airflow around the waist is faster, and heat dissipation is more even. (2) Differences in muscle activity: When standing, the waist muscles are slightly tense to maintain balance, and the local metabolic heat production may be slightly higher than when sitting. (3) Changes in clothing fit: When sitting, clothing adheres more closely to the skin due to compression, while when standing, the clothing droops due to gravity, resulting in different insulation effects. Influenced by these factors, the sudden release of heat from a seated to a standing position causes the waist to rapidly diffuse outwards, leading to a change in the vertical temperature gradient from low to high. Conversely, during the transition from a standing to a sitting position, the local temperature rises after the waist comes into contact with the seat, causing the temperature gradient to decrease.

[0091] Therefore, when an increase in the vertical temperature gradient is detected, the posture information is identified as "sitting posture → standing posture", and the air supply mode is switched to the person-blowing mode; when a decrease in the vertical temperature gradient is detected, the system is identified as "standing posture → sitting posture", and the air supply mode is switched to the person-avoiding mode.

[0092] This implementation method achieves posture switching detection without the need for additional posture sensors by analyzing the physical mechanism of temperature distribution changes in the waist region. Compared with solutions that rely solely on static image recognition, this method utilizes the abrupt changes in temperature gradients for judgment, exhibiting stronger anti-interference capabilities and maintaining high accuracy even under conditions of changing lighting and cluttered backgrounds. Linking posture detection with airflow mode allows the air conditioner to respond instantly to changes in human posture, improving comfort and reducing unnecessary direct cold airflow.

[0093] In the above embodiments, the controller 107 is further configured to control the swing angle of the transverse and longitudinal air guides based on the attitude information and the corrected distance information, including: Based on the centroid coordinates of the head region The ratio of the horizontal coordinate to the distance information is used to determine the horizontal azimuth of the human body based on the arctangent function of the ratio. , ; Vertical coordinates based on the centroid coordinates of the head region and the centroid coordinates of the waist region The difference between the vertical coordinates and the set scaling factor. The product of the two factors determines the vertical azimuth angle of the human body. , ; The swing angles of the horizontal and vertical air guides are controlled based on the horizontal and vertical azimuth angles of the human body.

[0094] Based on the above configuration, the controller 107 calculates the spatial orientation of the human body in the air conditioning supply coordinate system by calculating the relative positions of the human head and waist in the image coordinate system.

[0095] First, the controller 107 acquires the horizontal coordinates of the centroid of the head region and divides them by the fused distance information to obtain the horizontal position ratio. The arctangent function of this ratio is then used to calculate the horizontal azimuth angle of the human body, which reflects the amount of left-right offset of the human body relative to the front of the air conditioner.

[0096] In vertical positioning, the controller 107 calculates the difference between the vertical coordinates of the head region's centroid and the vertical coordinates of the waist region's centroid, and multiplies this difference by a set scaling factor to obtain the vertical azimuth angle of the human body. This angle reflects the relative positional change of the human body in the up-down direction.

[0097] Finally, the controller 107 uses the horizontal azimuth angle to control the swing angle of the horizontal air guide plate and the vertical azimuth angle to control the swing angle of the vertical air guide plate, so as to achieve precise pointing of the air supply direction and enable the air conditioner to dynamically adjust the air supply mode according to the real-time changes in the position of the human body.

[0098] This implementation achieves three-dimensional spatial positioning based on the centroid coordinates of human body parts obtained through thermal imaging, calculating the body's azimuth angle without the need for additional LiDAR or 3D cameras. Compared to solutions relying solely on single-axis angle control, this method simultaneously calculates both horizontal and vertical azimuth angles, making the airflow direction more closely match the user's actual position and improving the targeting and comfort of the airflow. Through real-time linkage with the air guide motor, the system can maintain stable tracking of the airflow direction as the user moves, thereby reducing airflow deviation and improving the user experience.

[0099] In some embodiments, the controller 107 is further configured to: The head centroid coordinates of the head region and the waist centroid coordinates of the waist region are identified. The displacement of the head centroid coordinates and waist centroid coordinates in multiple consecutive thermal imaging images is calculated. Based on the displacement, the head centroid coordinates and waist centroid coordinates of the next moment are predicted using a linear extrapolation model. The swing angles of the horizontal and vertical air guides at the next moment are controlled based on the predicted head and waist center of mass coordinates.

[0100] Based on the above configuration, using a linear extrapolation model, the controller 107 can calculate the centroid coordinates of the head and waist at the next moment based on historical displacement information, thus achieving short-term prediction of the human body's position. The predicted position data will be used to calculate the swing angles of the horizontal and vertical air guides in advance, so that the airflow direction can be adjusted in time as the human body moves.

[0101] Compared to adjusting airflow solely based on the current frame position, introducing centroid coordinate prediction reduces airflow deviation caused by position update delays, making it particularly suitable for scenarios where users move within an indoor environment. By pre-controlling the air deflector angle, the air conditioner can adjust the airflow direction the instant a person reaches the predicted position, improving user comfort and avoiding the abrupt feeling of direct cold air blowing.

[0102] In the above implementation, the linear extrapolation model can be replaced by time-series prediction methods such as Kalman filtering, particle filtering, and LSTM (Long Short-Term Memory Network) to improve the prediction accuracy under complex trajectories such as accelerated motion and sudden turns.

[0103] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can reside in the same processor; or the above modules can be located in different processors in any combination.

[0104] In addition, this application also provides an air conditioner, which is the same as the above embodiment and will not be repeated here. The difference is that the controller is used to implement the control method of the air conditioner to control the movement of the air guide plate.

[0105] Figure 10 This is a flowchart of the control method according to an embodiment of this application, with reference to... Figure 10 As shown, the control methods include: In the target recognition step S111, a thermal imaging image of the infrared detection area is acquired, thermal spot clustering is performed on the thermal imaging image, the target area and background area in the thermal imaging image are identified, and the head area and waist area are identified. In distance acquisition step S112, a first distance to the target region is calculated based on the maximum thermal gradient difference between the target region and the background region, and a weighted calculation is performed on the first distance based on a second distance to obtain distance information; and, In the air supply control step S113, based on the posture information of the target object matched in the head area and waist area, the swing angle of the horizontal air guide plate and the vertical air guide plate is controlled according to the posture information and the corrected distance information, and the air supply mode of the air conditioner is switched to the blowing mode or the avoiding mode.

[0106] Based on the above steps, this embodiment achieves all-weather, high-precision human body detection and ranging through the fusion of thermal imaging and millimeter-wave radar, maintaining a high recognition rate even in conditions of insufficient light, insignificant temperature differences, or partial obstruction. By utilizing the relative positional relationship between the head and waist to obtain posture information, the airflow mode can be quickly switched when the user's movements change (such as sitting down or standing up), improving comfort and reducing energy waste. The air guide vane control, combined with human body azimuth angle calculation and closed-loop motor adjustment, significantly improves the accuracy and response speed of the airflow direction, enabling the air conditioner to intelligently track or actively avoid drafts.

[0107] It should be noted that the steps shown in the above process or in the flowchart of the accompanying figures can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0108] in addition, Figure 11 This is a schematic diagram of the hardware structure of the controller 107 according to an embodiment of this application.

[0109] The controller 107 may include a processor 110 and a memory 111 storing computer program instructions.

[0110] Specifically, the processor 110 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0111] The memory 111 may include a large-capacity storage device for data or instructions. For example, and not limitingly, the memory 111 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 111 may include removable or non-removable (or fixed) media. Where appropriate, the memory 111 may be internal or external to a data processing device. In a particular embodiment, the memory 111 is non-volatile memory. In a particular embodiment, the memory 111 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random Access Memory (FPMDRAM), Extended Data Out Dynamic Random Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0112] The memory 111 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 110.

[0113] The processor 110 implements any of the control methods described in the above embodiments by reading and executing computer program instructions stored in the memory 111.

[0114] In some embodiments, the controller 107 may further include a communication interface 112 and a bus 113. For example, Figure 11 As shown, the processor 110, memory 111, and communication interface 112 are connected through bus 113 and complete communication with each other.

[0115] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0116] For ease of explanation, the above description has been provided in conjunction with specific embodiments. However, the above exemplary discussion is not intended to be exhaustive or to limit the embodiments to the specific forms disclosed above. Various modifications and variations can be obtained based on the above teachings. The selection and description of the above embodiments are for the purpose of better explaining the principles and practical applications, thereby enabling those skilled in the art to better utilize the embodiments and various different variations of the embodiments suitable for specific application considerations.

Claims

1. An air conditioner, characterized in that, include: The indoor unit has an air outlet on its casing; A thermal imaging device is installed on the air outlet to collect thermal imaging images of the infrared detection area. An air guide plate is disposed at the air outlet, and the air guide plate includes a horizontal air guide plate and a vertical air guide plate; A controller, mounted on the indoor unit, is electrically connected to the thermal imaging device, and is configured to: The thermal imaging device acquires thermal imaging images of the infrared detection area, performs hot spot clustering on the thermal imaging images, identifies the target area and background area in the thermal imaging images, and identifies the head area and waist area. The thermal imaging images contain the temperature values ​​of each pixel. Calculate the first distance of the target region based on the maximum thermal gradient difference between the target region and the background region, determine the distance information of the target region based on the first distance and perform correction to obtain the corrected distance information; Based on the posture information of the target object matched by the head and waist regions, the swing angle of the horizontal and vertical air guides is controlled according to the posture information and the corrected distance information, and the air supply mode of the air conditioner is switched to the blowing mode or the avoiding mode.

2. The air conditioner according to claim 1, characterized in that, Also includes: A radar sensor, installed on the air outlet, is used to detect a second distance to the target area; The step of determining and correcting the distance information of the target area based on the first distance includes: The thermal imaging confidence level is calculated based on the temperature difference between the target area and the background area, and the radar confidence level is determined based on the multipath interference index. The first distance and the second distance are weighted and normalized based on the thermal imaging confidence level and the radar confidence level to obtain the distance information.

3. The air conditioner according to claim 2, characterized in that, The step of determining and correcting the distance information of the target area based on the first distance also includes: Based on the distance information, the horizontal deflection angle and the vertical deflection angle of the thermal imaging device, feature factors are constructed. The feature factors are calculated based on the elastic network regression model to obtain the corrected distance information.

4. The air conditioner according to claim 1, characterized in that, The controller is also configured to: The head centroid coordinates of the head region and the waist centroid coordinates of the waist region are identified. The displacement of the head centroid coordinates and waist centroid coordinates in multiple consecutive thermal imaging images is calculated. Based on the displacement, the head centroid coordinates and waist centroid coordinates of the next moment are predicted using a linear extrapolation model. The swing angles of the transverse and longitudinal air guides at the next moment are controlled based on the predicted head and waist centroid coordinates.

5. The air conditioner according to claim 4, characterized in that, The step of performing hotspot clustering on the thermal imaging image to identify the target region, background region, and head and waist regions in the thermal imaging image further includes: Pixels with temperatures greater than the human metabolic temperature threshold in the thermal imaging image are extracted as high-temperature points. Based on the clustering algorithm DBSCAN, the high-temperature points are clustered to identify multiple hot spot regions as target regions. A minimum rectangular region containing the hot spot region is determined as the target bounding region. The width and height of the target bounding region are expanded outward by a specified multiple, and pixels with temperature values ​​greater than the noise temperature threshold in the expanded region are excluded to obtain the background region. Calculate the ellipse fit of the hot spot region. If the ellipse fit exceeds a set ellipse threshold and the temperature standard deviation of the hot plate region is lower than the temperature difference threshold, then mark the corresponding hot spot region as the head region. Calculate the vertical temperature gradient of the hot spot region. If there exists a continuous column of pixels whose vertical temperature is greater than the gradient threshold, then mark the corresponding hot spot region as the waist region.

6. The air conditioner according to claim 5, characterized in that, The calculation of the first distance of the target region based on the maximum thermal gradient difference between the target region and the background region further includes: The maximum thermal gradient difference is obtained by calculating the difference between the highest temperature value in the target area and the average temperature value in the background area. The maximum thermal gradient difference is exponentially multiplied by the thermal radiation attenuation coefficient. The ratio of the product of the exponentiation result and the focal length of the thermal imaging device to the product of the calibration constant of the thermal imaging device and the alignment deviation angle is taken as the first distance, where the alignment deviation angle is the pixel coordinate deviation angle between the centroid coordinates of the target area and the center of the optical axis.

7. The air conditioner according to claim 2, characterized in that, The step of calculating the thermal imaging confidence level based on the temperature difference between the target area and the background area further includes: The temperature difference between the target area and the background area is used as the signal amplitude, and the standard deviation of the temperature measurement noise is used as the noise amplitude. The signal amplitude and noise amplitude are calculated using the amplitude signal-to-noise ratio calculation model to obtain the temperature signal-to-noise ratio. The negative value of the temperature signal-to-noise ratio is reduced by a factor of two. An exponential function with the natural constant as the base is taken, and the difference between 1 and the exponential function is taken as the confidence level of thermal imaging.

8. The air conditioner according to claim 5, characterized in that, The controller is also configured to: The vertical temperature gradient of the waist region in consecutive frames is monitored. If the vertical temperature gradient increases, the posture information is identified as a change from sitting to standing, and the air conditioner's air supply mode is switched to the blowing mode. If the vertical temperature gradient decreases, the posture information is identified as a change from standing to sitting, and the air conditioner's air supply mode is switched to the human avoidance mode.

9. The air conditioner according to claim 8, characterized in that, The controller is also configured to: Based on the ratio of the horizontal coordinate of the centroid coordinate of the head region to the distance information, the horizontal azimuth angle of the human body is determined according to the arctangent function of the ratio. The vertical azimuth angle of the human body is determined by multiplying the difference between the vertical coordinates of the centroid coordinates of the head region and the vertical coordinates of the centroid coordinates of the waist region with a set scaling factor. The swing angles of the horizontal and vertical air guides are controlled based on the horizontal and vertical azimuth angles of the human body.

10. An air conditioner, characterized in that, include: The indoor unit has an air outlet on its casing; A thermal imaging device is installed on the air outlet to collect thermal imaging images of the infrared detection area. A radar sensor, installed on the air outlet, is used to detect a second distance to the target area; An air guide plate is disposed at the air outlet, and the air guide plate includes a horizontal air guide plate and a vertical air guide plate; A controller, mounted on the indoor unit, is electrically connected to the thermal imaging device. The controller is used to implement a control method for the air conditioner, controlling the movement of the air guide vane. The control method includes: The target recognition step involves acquiring a thermal imaging image of the infrared detection area, performing hot spot clustering on the thermal imaging image, identifying the target area and background area in the thermal imaging image, and identifying the head area and waist area. The distance acquisition step involves calculating a first distance to the target region based on the maximum thermal gradient difference between the target region and the background region, and then weighting the first distance based on the second distance to obtain distance information; and... The air supply control step involves matching the posture information of the target object based on the head and waist areas, controlling the swing angle of the horizontal and vertical air guides according to the posture information and the corrected distance information, and switching the air supply mode of the air conditioner to either the blowing mode or the avoiding mode.

Images