Air conditioner and hand region detection method

Abstract

The application discloses an air conditioner and a hand region detection method. A controller of the air conditioner is configured to generate an infrared image according to lumen data and indoor temperature; perform binaryzation processing on the infrared image according to at least two pixel threshold values and generate a binaryzation image corresponding to each pixel threshold value; split the binaryzation image in a vertical direction according to a preset splitting rule, and screen a target calibration region from each calibration region according to a splitting result; and determine a hand region according to a shape parameter of each target calibration region. The calibration region is a region with a continuous pixel value of a preset pixel value in the binaryzation image. The binaryzation processing based on the at least two pixel threshold values can extract hierarchical images of different temperature values, improve the adaptability and generalization ability to different infrared images, and exclude interference factors in combination with the shape parameter, thereby improving the accuracy of detecting the hand region without increasing the product cost.

Description

Technical Field

[0001] This application relates to the field of air conditioners, and more specifically, to a method for detecting air conditioners and hand areas. Background Technology

[0002] Because the human hand's sensitivity to hot and cold is far greater than that of other organs in the body, the temperature information of the hand can be used to immediately determine a person's true feelings about the external environment. Therefore, air conditioners can combine the temperature information of the hand to intelligently adjust the appropriate temperature and mode according to the body's temperature and cold sensation.

[0003] Due to concerns about the risk of privacy theft and leakage, the use of cameras in smart homes is largely unacceptable to users. Infrared sensors, on the other hand, have wide applications in home appliances. Many existing air conditioning products already use infrared sensors for related applications. However, due to cost and other issues, the resolution of the infrared sensors used is relatively low, unable to distinguish detailed features of human hands, with poor image contrast, and is greatly affected by hot air and other heat sources.

[0004] In addition, due to factors such as cost control of air conditioners, infrared sensors often use microcontrollers with weak computing power as the main control chip to carry out infrared sensor operations, thus they cannot use deep learning neural network methods that rely on hand feature information and chip computing power for hand detection.

[0005] Therefore, how to provide an air conditioner that can accurately detect the human hand area without increasing product costs is a technical problem that needs to be solved. Summary of the Invention

[0006] This invention provides an air conditioner to solve the technical problem of low accuracy when detecting hand areas using infrared sensors in the prior art.

[0007] The air conditioner includes:

[0008] Infrared sensors are used to detect indoor spaces and generate lumen data;

[0009] Indoor temperature sensor, used to detect indoor temperature;

[0010] The controller is configured as follows:

[0011] An infrared image is generated based on the lumen data and the indoor temperature;

[0012] The infrared image is binarized based on at least two pixel thresholds, and a binarized image corresponding to each pixel threshold is generated respectively.

[0013] The binarized image is segmented vertically according to a preset segmentation rule, and the target calibration region is selected from each calibration region based on the segmentation results.

[0014] The hand region is determined based on the shape parameters of each target calibration region;

[0015] The calibration region is the area in the binarized image where the pixel values ​​are continuously set to a preset pixel value.

[0016] In some embodiments of this application, the controller is specifically configured as follows:

[0017] The binarized image is divided into multiple sub-regions at equal intervals in the vertical direction;

[0018] Extract a middle region from the binarized image according to a preset cropping range;

[0019] The intermediate region is divided into multiple candidate regions according to a grid.

[0020] The calibration region located in each of the candidate regions is determined as the target calibration region;

[0021] The intermediate region is composed of one or more of the sub-regions, and the central position of the binarized image in the vertical direction is within the range of the intermediate region.

[0022] In some embodiments of this application, the number of sub-regions is 8, the number of candidate regions is 4, and the preset cutoff range is 3 / 8-5 / 8.

[0023] In some embodiments of this application, the controller is specifically configured as follows:

[0024] Valid target regions are selected from each of the target calibration regions based on the shape parameters;

[0025] The hand region is determined based on each of the effective target regions.

[0026] In some embodiments of this application, the shape parameters include region area, area ratio, and rectangle ratio, and the controller is further specifically configured to:

[0027] If the area of ​​the target calibration region is not greater than a preset area, and the area ratio of the target calibration region is not less than a preset area ratio, and the rectangle ratio of the target calibration region is greater than a first rectangle ratio and less than a second rectangle ratio, then the target calibration region is determined to be the effective target region; otherwise, the target calibration region is determined to be an interference object.

[0028] Wherein, the area of ​​the region is the number of pixels in the target calibration region, Q = sum / (H*W)*100, R = H / W*100, Q is the area ratio, sum is the area of ​​the region, R is the rectangle ratio, H is the height of the target calibration region, and W is the width of the target calibration region.

[0029] In some embodiments of this application, the controller is further specifically configured as follows:

[0030] Based on the coordinates of each effective target region, determine the intersection-union ratio between each effective target region and other effective target regions;

[0031] If the intersection-union ratio is greater than the preset percentage, the two effective target regions will be merged and determined as one hand region;

[0032] If the intersection-union ratio is not greater than the preset percentage, the two effective target areas are determined as two hand areas.

[0033] In some embodiments of this application, the controller is specifically configured as follows:

[0034] Convert the lumen data into temperature data according to Formula 1;

[0035] The temperature data is converted into pixel data according to Formula 2;

[0036] The infrared image is generated by representing the pixel data as a two-dimensional matrix.

[0037] Formula 1 is: T1 = (L - 8191) / 30 + T2.

[0038] Formula 2 is: I = (T1 - T) min ) / (T max -T min )*255,

[0039] T1 is the temperature data, L is the lumen data, T2 is the preset reference temperature of the infrared sensor control board, I is the pixel data, and T min The first preset temperature is defined as the indoor temperature when it is lower than the first preset temperature. min The second preset temperature is defined as the indoor temperature being greater than the second preset temperature. max The third preset temperature is the temperature at which the first preset temperature is lower than the second preset temperature, and the second preset temperature is lower than the third preset temperature.

[0040] In some embodiments of this application, the controller is specifically configured as follows:

[0041] Each pixel threshold is compared with the pixel value in the infrared image, and each binarized image is generated based on the comparison result.

[0042] Specifically, pixel values ​​in the infrared image that are greater than the pixel threshold are set to 255, and pixel values ​​in the infrared image that are not greater than the pixel threshold are set to 0.

[0043] In some embodiments of this application, the controller is further configured to:

[0044] Before binarizing the infrared image based on at least two pixel thresholds, the infrared image is denoised based on a median filtering algorithm, and the resolution of the infrared image is enlarged based on an interpolation algorithm.

[0045] After generating each of the binarized images, each of the binarized images is processed based on the image opening operation.

[0046] Accordingly, the present invention also proposes a method for detecting hand areas, applied in an air conditioner including an infrared sensor, an indoor temperature sensor, and a controller, the method comprising:

[0047] An infrared image is generated based on the lumen data obtained from the infrared sensor and the indoor temperature obtained from the indoor temperature sensor.

[0048] The infrared image is binarized based on at least two pixel thresholds, and a binarized image corresponding to each pixel threshold is generated respectively.

[0049] The binarized image is segmented vertically according to a preset segmentation rule, and the target calibration region is selected from each calibration region based on the segmentation results.

[0050] The hand region is determined based on the shape parameters of each target calibration region;

[0051] The calibration region is the area in the binarized image where the pixel values ​​are continuously set to a preset pixel value.

[0052] By applying the above technical solutions, in an air conditioner including an infrared sensor, an indoor temperature sensor, and a controller, the controller is configured to: generate an infrared image based on the lumen data and the indoor temperature; binarize the infrared image based on at least two pixel thresholds and generate binarized images corresponding to each pixel threshold; segment the binarized image vertically according to a preset segmentation rule, and select target calibration regions from each calibration region based on the segmentation results; determine the hand region based on the shape parameters of each target calibration region; wherein, the calibration region is the region in the binarized image where the pixel value is a preset pixel value. Binarization based on at least two pixel thresholds can extract layered images of different temperature values, improving the adaptability and generalization ability to different infrared images, and combining shape parameters to eliminate interference factors, thereby improving the accuracy of detecting the hand region without increasing product costs. Attached Figure Description

[0053] 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.

[0054] Figure 1 A schematic diagram of the structure of an air conditioner according to an embodiment of the present invention is shown;

[0055] Figure 2 This diagram illustrates the principle of determining the target calibration region in an embodiment of the present invention.

[0056] Figure 3 A flowchart illustrating a hand region detection method proposed in an embodiment of the present invention is shown;

[0057] Figure 4 A flowchart illustrating a hand region detection method according to another embodiment of the present invention is shown. Detailed Implementation

[0058] 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 some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0059] 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 technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0060] In this application, the air conditioner performs a refrigeration cycle by using a compressor, condenser, expansion valve, and evaporator. The refrigeration cycle includes a series of processes involving compression, condensation, expansion, and evaporation, and supplies refrigerant to the conditioned and heat-exchanged air.

[0061] The compressor compresses the refrigerant gas under high temperature and high pressure and discharges the compressed refrigerant gas. The discharged refrigerant gas flows into the condenser, where the condenser condenses the compressed refrigerant into a liquid phase, and the heat is released to the surrounding environment through the condensation process.

[0062] The expansion valve expands the high-temperature, high-pressure liquid refrigerant condensed in the condenser into a low-pressure liquid refrigerant. The evaporator evaporates the expanded refrigerant in the expansion valve, returning the low-temperature, low-pressure refrigerant gas to the compressor. The evaporator achieves its cooling effect by utilizing the latent heat of refrigerant evaporation to exchange heat with the material being cooled. Throughout the cycle, the air conditioner regulates the temperature of the indoor space.

[0063] The outdoor unit of an air conditioner refers to the part of the refrigeration cycle that includes the compressor and the outdoor heat exchanger. The indoor unit of an air conditioner includes the indoor heat exchanger, and an expansion valve can be provided in either the indoor or outdoor unit.

[0064] The indoor and outdoor heat exchangers function as either condensers or evaporators. When the indoor heat exchanger is used as a condenser, the air conditioner functions as a heater in heating mode; when the indoor heat exchanger is used as an evaporator, the air conditioner functions as a cooler in cooling mode.

[0065] This application provides an air conditioner, such as... Figure 1 As shown, it includes:

[0066] Infrared sensor 10 is used to detect the indoor space and output lumen data;

[0067] Indoor temperature sensor 20, used to detect indoor temperature;

[0068] Controller 30 is configured as follows:

[0069] An infrared image is generated based on the lumen data and the indoor temperature;

[0070] The infrared image is binarized based on at least two pixel thresholds, and a binarized image corresponding to each pixel threshold is generated respectively.

[0071] The binarized image is segmented vertically according to a preset segmentation rule, and the target calibration region is selected from each calibration region based on the segmentation results.

[0072] The hand region is determined based on the shape parameters of each target calibration region.

[0073] In this embodiment, the infrared sensor can be an infrared thermopile sensor, which is a sensor that uses infrared light for detection. The infrared sensor detects the indoor space and outputs lumen data including multiple lumen values. The lumen value is generally around seven or eight thousand. The lumen value is large in areas with high temperature and relatively small in areas with low temperature.

[0074] The controller connects to an infrared sensor and an indoor temperature sensor. The infrared sensor acquires lumen data, and the indoor temperature sensor acquires indoor temperature data. Based on the lumen data and indoor temperature, an infrared image is generated. The infrared image is a single-channel image, i.e., a two-dimensional data set with values ​​between 0 and 255, similar to a grayscale image or a black and white photograph. Then, the infrared image is binarized according to at least two pixel thresholds, generating binarized images corresponding to each threshold. A binarized image is an image with pixel values ​​of 0 or 255; only these two values ​​exist, 0 representing pure black and 255 representing pure white. The process of converting a grayscale image to a binarized image is called binarization. Finally, the region in the binarized image where the pixel value is continuously at a preset value is defined as a calibration region. It can be understood that the number of binarized images is consistent with the number of pixel thresholds.

[0075] Since the hand occupies a small proportion of the human body, the binarized image needs to be segmented vertically according to the preset segmentation rules. Based on the segmentation results, the target calibration region can be selected from each calibration region. The preset segmentation rules can be determined based on the positional features of the hand on the human body. Finally, the hand region can be determined based on the shape parameters of each target calibration region.

[0076] Through the above process, the air conditioner can accurately detect the hand area. Optionally, the controller can also be configured to:

[0077] The target temperature and operating mode of the air conditioner are controlled based on the temperature of the hand area.

[0078] In order to reliably filter out the target calibration area, in some embodiments of this application, the controller is specifically configured as follows:

[0079] The binarized image is divided into multiple sub-regions at equal intervals in the vertical direction;

[0080] Extract a middle region from the binarized image according to a preset cropping range;

[0081] The intermediate region is divided into multiple candidate regions according to a grid.

[0082] The calibration region located in each of the candidate regions is determined as the target calibration region;

[0083] The intermediate region is composed of one or more of the sub-regions, and the central position of the binarized image in the vertical direction is within the range of the intermediate region.

[0084] In this embodiment, the binarized image is regarded as a complete human body region. First, the binarized image is divided into multiple sub-regions at equal distances in the vertical direction. Since the human hand region is generally located in the middle position in the height direction of the human body, a middle region is extracted from the binarized image according to a preset cropping range. This middle region is composed of one or more sub-regions and the middle position of the binarized image in the vertical direction is within the range of the middle region. Then, the middle region is divided into multiple candidate regions according to the grid. Finally, the calibration region located in each candidate region is determined as the target calibration region.

[0085] To improve the accuracy of hand region detection, in some embodiments of this application, the number of sub-regions is 8, the number of candidate regions is 4, and the preset cutoff range is 3 / 8-5 / 8.

[0086] In this embodiment, as Figure 2 As shown, the binarized image is divided into 8 sub-regions at equal intervals in the vertical direction. The region between 3 / 8 and 5 / 8 is taken as the middle region, and then the middle region is divided into 4 candidate regions.

[0087] To accurately determine the hand area, in some embodiments of this application, the controller is specifically configured as follows:

[0088] Valid target regions are selected from each of the target calibration regions based on the shape parameters;

[0089] The hand region is determined based on each of the effective target regions.

[0090] In this embodiment, the presence of a hand in the target calibration area is determined based on the shape parameters. If a hand is present, the target calibration area is a valid target area. Valid target areas are then selected from the target calibration areas, and the hand area can be determined based on each valid target area.

[0091] To accurately determine the effective target area, in some embodiments of this application, the shape parameters include the area, area ratio, and rectangle ratio, and the controller is further specifically configured as follows:

[0092] If the area of ​​the target calibration region is not greater than a preset area, and the area ratio of the target calibration region is not less than a preset area ratio, and the rectangle ratio of the target calibration region is greater than a first rectangle ratio and less than a second rectangle ratio, then the target calibration region is determined to be the valid target region; otherwise, the target calibration region is determined to be an interference object.

[0093] In this embodiment, the area is the number of pixels in the target calibration area, Q = sum / (H*W)*100, R = H / W*100, Q is the area ratio, sum is the area, R is the rectangle ratio, H is the height of the target calibration area, and W is the width of the target calibration area.

[0094] Optionally, the preset area is 240, the preset area ratio is 50, the first rectangle ratio is 50, and the second rectangle ratio is 150.

[0095] Those skilled in the art may select other shape parameters or increase or decrease the number of shape parameters according to actual needs, which does not affect the scope of protection of this application.

[0096] Since different effective target areas may overlap, in order to accurately determine the hand area, in some embodiments of this application, the controller is further configured to:

[0097] Based on the coordinates of each effective target region, determine the intersection-union ratio between each effective target region and other effective target regions;

[0098] If the intersection-union ratio is greater than the preset percentage, the two effective target regions will be merged and determined as one hand region;

[0099] If the intersection-union ratio is not greater than the preset percentage, the two effective target areas are determined as two hand areas.

[0100] In this embodiment, the Intersection over Union (IOU) ratio of regions A and B is calculated using the formula: IOU = (A∩B) / (A∪B). The coordinates of each effective target region represent its position in each binarized image. Based on these coordinates, the IOU ratio between each effective target region and other effective target regions is determined. The IOU ratio is the ratio of the intersection to the union of two effective target regions. If the IOU ratio is greater than a preset percentage, the two effective target regions are merged into one hand region; if the IOU ratio is not greater than the preset percentage, the two effective target regions are determined as two hand regions.

[0101] It should be noted that the above embodiments are only one specific implementation scheme proposed in this application, and other methods of determining the hand area based on each effective target area are all within the protection scope of this application.

[0102] In order to accurately generate infrared images, in some embodiments of this application, the controller is specifically configured as follows:

[0103] Convert the lumen data into temperature data according to Formula 1;

[0104] The temperature data is converted into pixel data according to Formula 2;

[0105] The infrared image is generated by representing the pixel data as a two-dimensional matrix.

[0106] Formula 1 is: T1 = (L - 8191) / 30 + T2.

[0107] Formula 2 is: I = (T1 - T) min ) / (T max -T min )*255,

[0108] T1 is the temperature data, L is the lumen data, T2 is the preset reference temperature of the infrared sensor control board, I is the pixel data, and T min The first preset temperature is defined as the indoor temperature when it is lower than the first preset temperature. min The second preset temperature is defined as the indoor temperature being greater than the second preset temperature. max The third preset temperature is the temperature at which the first preset temperature is lower than the second preset temperature, and the second preset temperature is lower than the third preset temperature.

[0109] Optionally, the first preset temperature is 20℃, the second preset temperature is 32℃, and the third preset temperature is 38℃.

[0110] It should be noted that the above embodiments are only one specific implementation scheme proposed in this application, and other methods of generating infrared images based on lumen data and indoor temperature are all within the protection scope of this application.

[0111] In order to accurately generate binarized images, in some embodiments of this application, the controller is specifically configured as follows:

[0112] Each pixel threshold is compared with the pixel value in the infrared image, and each binarized image is generated based on the comparison result.

[0113] Specifically, pixel values ​​in the infrared image that are greater than the pixel threshold are set to 255, and pixel values ​​in the infrared image that are not greater than the pixel threshold are set to 0.

[0114] For example, if the pixel threshold is 128, the pixel values ​​in the infrared image greater than 128 are set to 255, and the pixel values ​​in the infrared image not greater than 128 are set to 0. The preset pixel value can be 255.

[0115] To improve the accuracy of hand area detection, in some embodiments of this application, the controller is further configured to:

[0116] Before binarizing the infrared image based on at least two pixel thresholds, the infrared image is denoised based on a median filtering algorithm, and the resolution of the infrared image is enlarged based on an interpolation algorithm.

[0117] After generating each of the binarized images, each of the binarized images is processed based on the image opening operation.

[0118] In this embodiment, the infrared image is denoised using a median filtering algorithm, which reduces image noise and preserves as many details as possible. The resolution of the infrared image is increased using an interpolation algorithm, which increases the image data and makes the hand target more prominent and accurate.

[0119] Median filtering can be a 3x3 median filter, which sorts the nine values ​​in the 3x3 matrix and outputs the median value. The interpolation algorithm can be a bicubic interpolation algorithm, which increases the image resolution by a factor of four, changing the image size from 80*60 to 320*240.

[0120] Those skilled in the art can flexibly choose different median filtering algorithms and interpolation algorithms according to actual needs, which does not affect the scope of protection of this application. In addition, the specific process of image processing based on median filtering algorithms and interpolation algorithms is prior art and will not be described in detail here.

[0121] Image opening is a common technique in morphological processing. Opening operations typically smooth object contours, break narrow necks, and eliminate fine protrusions. However, the unique way the fingers of a hand are spread, especially the tiny connection between the thumb and palm, can severely affect the accuracy of parameters in the binarized hand region, thus impacting the detection results. To address this issue, we combine morphological processing with opening operations on the binarized image, specifically erosion followed by dilation. This removes noise, eliminates small objects, separates objects at thin points, and smooths the boundaries of larger objects without significantly altering their area.

[0122] By applying the above technical solutions, in an air conditioner including an infrared sensor, an indoor temperature sensor, and a controller, the controller is configured to: generate an infrared image based on the lumen data and the indoor temperature; binarize the infrared image based on at least two pixel thresholds and generate binarized images corresponding to each pixel threshold; segment the binarized image vertically according to a preset segmentation rule, and select target calibration regions from each calibration region based on the segmentation results; determine the hand region based on the shape parameters of each target calibration region; wherein, the calibration region is the region in the binarized image where the pixel value is a preset pixel value. Binarization based on at least two pixel thresholds can extract layered images of different temperature values, improving the adaptability and generalization ability to different infrared images, and combining shape parameters to eliminate interference factors, thereby improving the accuracy of detecting the hand region without increasing product costs.

[0123] To further illustrate the technical concept of this invention, the technical solution of this invention will now be described in conjunction with specific application scenarios.

[0124] This application provides a method for detecting a hand area, applied in an air conditioner including an infrared sensor, an indoor temperature sensor, and a controller, such as... Figure 3 As shown, the method includes the following steps:

[0125] Step S101, Begin.

[0126] Step S102: Generate an infrared image.

[0127] In this step, an infrared image is generated based on lumen data and indoor temperature.

[0128] Step S103, image preprocessing.

[0129] In this step, the infrared image is denoised using a median filtering algorithm, and its resolution is increased using an interpolation algorithm.

[0130] Step S104: Generate a binarized image.

[0131] Step S105: Is there a binarized image under the current pixel threshold? If yes, proceed to step S106; otherwise, proceed to step S110.

[0132] Step S106, image opening operation.

[0133] In this step, each binarized image is processed based on image opening operations.

[0134] Step S107: Segment the binarized image and filter out the target calibration region.

[0135] Specifically, the binarized image is divided into 8 sub-regions at equal intervals in the vertical direction. The region between 3 / 8 and 5 / 8 is taken as the middle region. The middle region is then divided into 4 candidate regions. The calibration region in each candidate region is determined as the target calibration region.

[0136] Step S108: Is the target calibration area a valid target area? If yes, proceed to step S109; otherwise, proceed to step S110.

[0137] Valid target areas are selected from each target calibration area based on the area, area ratio, and rectangle ratio.

[0138] Step S109: Save the current valid target area.

[0139] Step S110: All pixel thresholds have been traversed. If so, proceed to step S111; otherwise, proceed to step S104.

[0140] Step S111: Integrate all valid target areas.

[0141] In this step, if the intersection-union ratio is greater than the preset percentage, the two valid target areas are merged into one hand area; if the intersection-union ratio is not greater than the preset percentage, the two valid target areas are determined as two hand areas.

[0142] Step S112: Output the hand area.

[0143] Step S113, End.

[0144] This application also proposes a method for detecting hand areas, applied in an air conditioner including an infrared sensor, an indoor temperature sensor, and a controller, such as... Figure 4 As shown, the method includes the following steps:

[0145] Step S201: Generate an infrared image based on the lumen data obtained from the infrared sensor and the indoor temperature obtained from the indoor temperature sensor;

[0146] Step S202: Binarize the infrared image according to at least two pixel thresholds and generate a binarized image corresponding to each pixel threshold.

[0147] Step S203: The binarized image is segmented vertically according to a preset segmentation rule, and the target calibration region is selected from each calibration region based on the segmentation result.

[0148] Step S204: Determine the hand region based on the shape parameters of each target calibration region;

[0149] The calibration region is the area in the binarized image where the pixel values ​​are continuously set to a preset pixel value.

[0150] 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 of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An air conditioner, characterized in that, include: Infrared sensors are used to detect indoor spaces and generate lumen data; Indoor temperature sensor, used to detect indoor temperature; The controller is configured as follows: An infrared image is generated based on the lumen data and the indoor temperature; The infrared image is binarized based on at least two pixel thresholds, and a binarized image corresponding to each pixel threshold is generated respectively. The binarized image is segmented vertically according to a preset segmentation rule, and the target calibration region is selected from each calibration region based on the segmentation results. The hand region is determined based on the shape parameters of each target calibration region; The calibration region is the region in the binarized image where the pixel values ​​are continuously equal to a preset pixel value; The controller is specifically configured as follows: The binarized image is divided into multiple sub-regions at equal intervals in the vertical direction; a middle region is extracted from the binarized image according to a preset cropping range; the middle region is divided into multiple candidate regions according to a grid; the calibration region in each of the candidate regions is determined as the target calibration region; wherein, the middle region is composed of one or more of the sub-regions, and the middle position of the binarized image in the vertical direction is within the range of the middle region; Valid target regions are selected from each of the target calibration regions based on the shape parameters. The shape parameters include region area, area ratio, and rectangle ratio. If the region area of ​​the target calibration region is not greater than a preset area, and the area ratio of the target calibration region is not less than a preset area ratio, and the rectangle ratio of the target calibration region is greater than a first rectangle ratio and less than a second rectangle ratio, the target calibration region is determined to be a valid target region; otherwise, the target calibration region is determined to be an interference object. Wherein, the region area is the number of pixels in the target calibration region, Q = sum / (H*W)*100, R = H / W*100, Q is the area ratio, sum is the region area, R is the rectangle ratio, H is the height of the target calibration region, and W is the width of the target calibration region. The hand region is determined based on each of the effective target regions.

2. The air conditioner as described in claim 1, characterized in that, The number of sub-regions is 8, the number of candidate regions is 4, and the preset interception range is 3 / 8-5 / 8.

3. The air conditioner as described in claim 1, characterized in that, The controller is further specifically configured as follows: Based on the coordinates of each effective target region, determine the intersection-union ratio between each effective target region and other effective target regions; If the intersection-union ratio is greater than the preset percentage, the two effective target regions will be merged and determined as one hand region; If the intersection-union ratio is not greater than the preset percentage, the two effective target areas are determined as two hand areas.

4. The air conditioner as described in claim 1, characterized in that, The controller is specifically configured as follows: Convert the lumen data into temperature data according to Formula 1; The temperature data is converted into pixel data according to Formula 2; The infrared image is generated by representing the pixel data as a two-dimensional matrix. Formula 1 is: T1 = (L - 8191) / 30 + T2. Formula 2 is: I = (T1 - T) min ) / (T max -T min )*255, T1 is the temperature data, L is the lumen data, T2 is the preset reference temperature of the infrared sensor control board, I is the pixel data, and T min The first preset temperature is defined as the indoor temperature when it is lower than the first preset temperature. min The second preset temperature is defined as the indoor temperature being greater than the second preset temperature. max The third preset temperature is the temperature at which the first preset temperature is lower than the second preset temperature, and the second preset temperature is lower than the third preset temperature.

5. The air conditioner as described in claim 1, characterized in that, The controller is specifically configured as follows: Each pixel threshold is compared with the pixel value in the infrared image, and each binarized image is generated based on the comparison result. Specifically, pixel values ​​in the infrared image that are greater than the pixel threshold are set to 255, and pixel values ​​in the infrared image that are not greater than the pixel threshold are set to 0.

6. The air conditioner as described in claim 1, characterized in that, The controller is also configured to: Before binarizing the infrared image based on at least two pixel thresholds, the infrared image is denoised based on a median filtering algorithm, and the resolution of the infrared image is enlarged based on an interpolation algorithm. After generating each of the binarized images, each of the binarized images is processed based on the image opening operation.

7. A method for detecting a hand region, characterized in that, The method, applied in an air conditioner including an infrared sensor, an indoor temperature sensor, and a controller, comprises: An infrared image is generated based on the lumen data obtained from the infrared sensor and the indoor temperature obtained from the indoor temperature sensor. The infrared image is binarized based on at least two pixel thresholds, and a binarized image corresponding to each pixel threshold is generated respectively. The binarized image is segmented vertically according to a preset segmentation rule, and the target calibration region is selected from each calibration region based on the segmentation results. The hand region is determined based on the shape parameters of each target calibration region; The calibration region is the region in the binarized image where the pixel values ​​are continuously equal to a preset pixel value; The controller is specifically configured as follows: The binarized image is divided into multiple sub-regions at equal intervals in the vertical direction; a middle region is extracted from the binarized image according to a preset cropping range; the middle region is divided into multiple candidate regions according to a grid; the calibration region in each of the candidate regions is determined as the target calibration region; wherein, the middle region is composed of one or more of the sub-regions, and the middle position of the binarized image in the vertical direction is within the range of the middle region; Valid target regions are selected from each of the target calibration regions based on the shape parameters. The shape parameters include region area, area ratio, and rectangle ratio. If the region area of ​​the target calibration region is not greater than a preset area, and the area ratio of the target calibration region is not less than a preset area ratio, and the rectangle ratio of the target calibration region is greater than a first rectangle ratio and less than a second rectangle ratio, the target calibration region is determined to be a valid target region; otherwise, the target calibration region is determined to be an interference object. Wherein, the region area is the number of pixels in the target calibration region, Q = sum / (H*W)*100, R = H / W*100, Q is the area ratio, sum is the region area, R is the rectangle ratio, H is the height of the target calibration region, and W is the width of the target calibration region. The hand region is determined based on each of the effective target regions.

Images