Range hood control method and device based on gesture recognition, range hood and electronic equipment

Abstract

The invention discloses a range hood control method and device based on gesture recognition, a range hood and electronic equipment, and belongs to the technical field of kitchen appliances. The method comprises the following steps: controlling a laser transmitter to transmit a laser signal to a target direction of the range hood, and receiving a reflected light signal formed by reflecting the laser signal through a hand of a user through a single-photon avalanche diode; generating a depth map including the hand of the user according to the time difference between the laser signal and the reflected light signal; identifying a user gesture according to the depth map; and generating a control instruction corresponding to the user gesture to control the running state of the range hood. According to the embodiment of the invention, the single photon avalanche diode and the laser transmitter are arranged on the range hood, so that non-contact control of the range hood can be realized, the problem that oil stains on the hand of a user pollute components of the range hood in the cooking process is reduced, and the operation convenience of the range hood is improved.

Description

Technical Field

[0001] This application belongs to the field of kitchen appliance technology, and in particular relates to a range hood control method, device, range hood and electronic equipment based on gesture recognition. Background Technology

[0002] The core function of a range hood is to generate negative pressure through an internal fan system, capture oil fumes, gas exhaust gases, and suspended particulate matter produced during cooking, and then exhaust them outdoors through exhaust pipes, thereby improving kitchen air quality.

[0003] In related technologies, range hoods generally adopt a contact-based human-machine interaction solution, that is, buttons are set on the front or top surface of the casing, and users can control the range hood by directly pressing or touching the buttons with their fingers. However, during the cooking process, users' hands are usually contaminated with oil, water stains, flour and other stains, and users usually need to clean their hands before controlling the range hood by touch, which brings inconvenience to the user. Summary of the Invention

[0004] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes a method, device, range hood, and electronic device for controlling a range hood based on gesture recognition, so as to realize non-contact control of the range hood and improve the convenience of range hood operation.

[0005] In a first aspect, this application provides a gesture recognition-based method for controlling a tobacco appliance, wherein the tobacco appliance is equipped with a single-photon avalanche diode and a laser emitter; the method includes: The laser emitter is controlled to emit a laser signal toward the target direction of the smoke machine, and the reflected light signal formed by the laser signal after being reflected by the user's hand is received by a single-photon avalanche diode. A depth map, including the user's hand, is generated based on the time difference between the laser signal and the reflected light signal. Recognize user gestures based on depth maps; Generate control commands corresponding to user gestures to control the operation of the range hood.

[0006] According to the gesture recognition-based range hood control method of this application, a laser emitter is controlled to emit a laser signal in the target direction of the range hood, and a single-photon avalanche diode is used to receive the reflected light signal formed by the laser signal after reflection from the user's hand. A depth map including the user's hand is generated based on the time difference between the laser signal and the reflected light signal. The user's gesture is recognized based on the depth map. A control command corresponding to the user's gesture is generated to control the operating state of the range hood. This embodiment of the application configures a single-photon avalanche diode and a laser emitter on the range hood. The laser emitter emits a laser signal in the target direction, and the single-photon avalanche diode captures the reflected light signal formed by reflection from the user's hand. Based on the time difference between the laser signal and the reflected light signal, a depth map reflecting the spatial position and morphological characteristics of the user's hand can be generated. Thus, the user's gesture can be recognized based on the depth map, and a control command corresponding to the recognition result can be generated to regulate the operating state of the range hood. This achieves non-contact control of the range hood, reduces the problem of oil contamination of the range hood components by the user's hands during cooking, and improves the convenience of operating the range hood.

[0007] According to one embodiment of this application, the single-photon avalanche diode is an array-type single-photon avalanche diode, which corresponds to multiple pixels, and each pixel corresponds to an independent single-photon avalanche diode; the method further includes: When the laser emitter emits a laser signal toward the target direction of the smoke machine, the emission time is recorded by the pixel timing circuit; When a laser signal is received by a single-photon avalanche diode and reflected by the user's hand, the receiving time is recorded by the timing circuit of the corresponding pixel of the single-photon avalanche diode. The time difference between the laser signal and the reflected light signal is calculated based on the transmission and reception times.

[0008] In this embodiment, by configuring an array of single-photon avalanche diodes and a laser emitter on the tobacco machine, with each array of single-photon avalanche diodes having multiple pixels and each pixel having an independent single-photon avalanche diode and timing circuit, the timing circuit of each pixel can synchronously record the laser emission time when the laser emitter emits a laser signal toward the target direction of the tobacco machine. When the array of single-photon avalanche diodes receives the reflected light signal formed by the user's hand, the timing circuit of the corresponding pixel can capture the received time of the reflected light, thereby improving the accuracy of time difference calculation.

[0009] According to one embodiment of this application, generating a depth map including a user's hand based on the time difference between a laser signal and a reflected light signal includes: Point cloud data representing the user's hand position in three-dimensional space is calculated based on the time difference between the laser signal and the reflected light signal, resulting in a depth map including the user's hand.

[0010] In this embodiment, point cloud data representing the three-dimensional spatial position of the user's hand is calculated by the time difference between the laser signal and the reflected light signal. The point cloud data can represent the three-dimensional shape, spatial coordinates and contour details of the user's hand in three-dimensional space, so that the generated depth map can accurately reflect the relative positional relationship and spatial distribution characteristics of various parts of the hand.

[0011] According to one embodiment of this application, point cloud data characterizing the three-dimensional spatial position of a user's hand is calculated based on the time difference between the laser signal and the reflected light signal, including: The distance between the user's hand and the single-photon avalanche diode is calculated based on the time difference between the laser signal and the reflected light signal. Based on the position information and distance of the pixels corresponding to the single-photon avalanche diode, point cloud data representing the position of the user's hand in three-dimensional space is generated.

[0012] In this embodiment, point cloud data is generated by calculating the distance between each position of the user's hand and the single-photon avalanche diode, and then combining the pixel position information. This provides radial depth and spatial coordinate references for the generation of point cloud data, realizing the mapping of three-dimensional coordinates of each position of the hand, so that the generated point cloud data can accurately reproduce the three-dimensional structure, contour details and spatial posture of the hand.

[0013] According to one embodiment of this application, recognizing user gestures based on depth maps includes: Obtain depth maps at multiple consecutive time points; Identify the user's hand movement trajectory based on depth maps from multiple consecutive time points; The user's gestures are determined based on the movement trajectory.

[0014] In this embodiment, by acquiring depth maps at multiple consecutive moments, it is possible to capture the spatial position and shape changes of the user's hand at different points in time, reconstruct the dynamic change process of the hand from a time dimension, thereby improving the accuracy of recognizing the user's hand movement trajectory and thus improving the accuracy of the user's gestures.

[0015] According to one embodiment of this application, user gestures are identified based on depth maps to obtain user gestures, including: Obtain depth maps at multiple consecutive time points; The depth maps at multiple consecutive time points are input into the gesture recognition model to obtain the user gestures output by the gesture recognition model.

[0016] In this embodiment, by acquiring depth maps at multiple consecutive moments, the spatial position and shape changes of the user's hand at different points in time can be captured, and the dynamic change process of the hand can be reconstructed from the time dimension. The continuous depth maps are input into the gesture recognition model, and the gesture recognition model can extract high-order features from the depth maps through deep learning algorithms, thereby accurately recognizing the user's gestures.

[0017] According to one embodiment of this application, User gestures include at least one of waving, hovering, clicking, swiping, and drawing circles; control commands include at least one of power on, power off, gear adjustment, light switch, and delayed power off.

[0018] Secondly, this application provides a gesture recognition-based control device for a tobacco appliance, wherein the tobacco appliance is equipped with a single-photon avalanche diode and a laser emitter; the device includes: The transmitting and receiving module is used to control the laser transmitter to emit laser signals in the target direction of the smoke machine, and to receive the reflected light signal formed by the laser signal after being reflected by the user's hand through a single-photon avalanche diode; The generation module is used to generate a depth map, including the user's hand, based on the time difference between the laser signal and the reflected light signal. The recognition module is used to recognize user gestures based on the depth map; The control module is used to generate control commands corresponding to user gestures to control the operating status of the range hood.

[0019] According to the gesture recognition-based range hood control device of this application, a laser emitter is controlled to emit a laser signal in the target direction of the range hood, and a single-photon avalanche diode is used to receive the reflected light signal formed by the laser signal after reflection from the user's hand. A depth map including the user's hand is generated based on the time difference between the laser signal and the reflected light signal. The user's gesture is recognized based on the depth map. A control command corresponding to the user's gesture is generated to control the operating state of the range hood. The embodiment of this application configures a single-photon avalanche diode and a laser emitter on the range hood. The laser emitter emits a laser signal in the target direction, and the single-photon avalanche diode captures the reflected light signal formed by reflection from the user's hand. Based on the time difference between the laser signal and the reflected light signal, a depth map reflecting the spatial position and shape characteristics of the user's hand can be generated. Thus, the user's gesture can be recognized based on the depth map, and a control command corresponding to the recognition result can be generated to regulate the operating state of the range hood. This achieves non-contact control of the range hood, reduces the problem of oil and grease from the user's hands contaminating the range hood components during cooking, and improves the convenience of operating the range hood.

[0020] Thirdly, this application provides a smoke machine, including a single-photon avalanche diode, a laser emitter, and a controller; A controller for performing the gesture recognition-based range hood control method as described in the first aspect above.

[0021] Fourthly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the gesture recognition-based smoke hood control method as described in the first aspect above.

[0022] Fifthly, this application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the gesture recognition-based smoke machine control method as described in the first aspect above.

[0023] The above-described technical solutions in this application embodiment control a laser emitter to emit a laser signal towards the target direction of the range hood, and receive the reflected light signal formed by the laser signal reflected by the user's hand through a single-photon avalanche diode; generate a depth map including the user's hand based on the time difference between the laser signal and the reflected light signal; recognize the user's gesture based on the depth map; and generate control commands corresponding to the user's gesture to control the operating state of the range hood. This application embodiment, by configuring a single-photon avalanche diode and a laser emitter on the range hood, uses the laser emitter to emit a laser signal towards the target direction and uses the single-photon avalanche diode to capture the reflected light signal formed by the user's hand. Based on the time difference between the laser signal and the reflected light signal, a depth map reflecting the spatial position and morphological characteristics of the user's hand can be generated. Thus, the user's gesture can be recognized based on the depth map, and control commands corresponding to the recognition result can be generated to regulate the operating state of the range hood. This achieves non-contact control of the range hood, reduces the problem of oil contamination of the range hood components by the user's hands during cooking, and improves the convenience of operating the range hood.

[0024] In some embodiments, by configuring an array of single-photon avalanche diodes and a laser emitter on the smoke machine, with the array of single-photon avalanche diodes corresponding to multiple pixels and each pixel configured with an independent single-photon avalanche diode and timing circuit, the timing circuit of each pixel can synchronously record the laser emission time when the laser emitter emits a laser signal toward the target direction of the smoke machine. When the array of single-photon avalanche diodes receives the reflected light signal formed by the user's hand, the timing circuit of the corresponding pixel can capture the received time of the reflected light, thereby improving the accuracy of time difference calculation.

[0025] In some embodiments, point cloud data representing the three-dimensional spatial position of the user's hand is calculated by the time difference between the laser signal and the reflected light signal. The point cloud data can represent the three-dimensional shape, spatial coordinates and contour details of the user's hand in three-dimensional space, so that the generated depth map can accurately reflect the relative positional relationship and spatial distribution characteristics of various parts of the hand.

[0026] In some embodiments, point cloud data is generated by calculating the distance between each position of the user's hand and the single-photon avalanche diode, and then combining the pixel position information. This provides radial depth and spatial coordinate references for the generation of point cloud data, realizes the mapping of three-dimensional coordinates of each position of the hand, and enables the generated point cloud data to accurately reproduce the three-dimensional structure, contour details and spatial posture of the hand.

[0027] In some embodiments, by acquiring depth maps at multiple consecutive moments, it is possible to capture the spatial position and shape changes of the user's hand at different points in time, reconstruct the dynamic change process of the hand from a time dimension, thereby improving the accuracy of recognizing the motion trajectory of the user's hand and thus improving the accuracy of the user's gestures.

[0028] In some embodiments, by acquiring depth maps at multiple consecutive moments, it is possible to capture the spatial position and shape changes of the user's hand at different points in time, reconstruct the dynamic change process of the hand from the time dimension, and input the continuous depth maps into the gesture recognition model. The gesture recognition model can extract high-order features from the depth maps through deep learning algorithms, thereby accurately recognizing the user's gestures.

[0029] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0030] The above and / or additional aspects and advantages of this application will become apparent and readily understood in conjunction with the following description of the embodiments in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart illustrating the gesture recognition-based range hood control method provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of the gesture recognition-based smoke hood control device provided in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0031] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0032] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0033] The following description, in conjunction with the accompanying drawings, details the gesture recognition-based range hood control method, device, range hood, and electronic equipment provided in this application through specific embodiments and application scenarios.

[0034] The gesture recognition-based range hood control method provided in this application embodiment can be executed by an electronic device or a functional module or entity in an electronic device that can implement the gesture recognition-based range hood control method. For example, the electronic devices mentioned in this application embodiment include, but are not limited to, range hood controllers, servers, etc. The following uses an electronic device as the execution subject to illustrate the gesture recognition-based range hood control method provided in this application embodiment.

[0035] A range hood, also known as a kitchen exhaust hood, is an essential piece of equipment in the kitchen used to remove cooking fumes and exhaust gases, maintaining fresh air in the kitchen. A range hood can include components such as a fan system, exhaust duct, control panel, air inlet, and air outlet. The air inlet is the part of the range hood that absorbs cooking fumes and exhaust gases; it is usually located at the bottom or side of the range hood, close to the stove, to capture the fumes generated during cooking. The air outlet is located at the top or rear of the range hood, expelling the absorbed fumes and exhaust gases outdoors through the exhaust duct. The fan system, as the core power component of the range hood, consists of a motor and an impeller. When the motor drives the impeller to rotate at high speed, a negative pressure area is created inside the fan. This pressure difference draws the cooking fumes into the machine and then exhausts them outdoors through the air outlet and exhaust duct. The control panel allows users to interact with the range hood. Users can control the range hood's on / off function, fan speed adjustment, lighting, and other functions through physical buttons or a touch interface on the control panel. In this embodiment, users can also control the range hood's on / off function, fan speed adjustment, lighting, and other functions through gestures. like Figure 1 As shown, the gesture recognition-based range hood control method of this application includes: step 110, step 120, step 130, and step 140.

[0036] Step 110: Control the laser emitter to emit a laser signal in the target direction of the smoke machine, and receive the reflected light signal formed by the laser signal after being reflected by the user's hand through a single-photon avalanche diode.

[0037] In this embodiment, the range hood may include a single-photon avalanche diode and a laser emitter. The laser emitter is a device that generates laser signals, such as a vertical-cavity surface-emitting laser (VCSEL). The laser emitter can emit laser pulses at a pulse repetition frequency under the modulation of the driving circuit. It should be noted that the peak pulse power of the laser emitter has undergone safety assessment and meets human eye safety standards, ensuring that there is no optical radiation hazard to users in the daily application scenarios of the range hood.

[0038] A single-photon avalanche diode (SPAD) is a semiconductor optoelectronic device with extremely high sensitivity to optical signals, enabling the detection of a single photon.

[0039] In some embodiments, the laser emitter and the single-photon avalanche diode can be arranged coplanarly or nearly coplanarly behind the same optical window on the inner side of the front panel of the smoke hood. The optical window can be closed by a narrow-band filter with high transmittance and low transmittance in the visible light band to suppress ambient light interference. Specifically, the laser emitter is located 10mm–20mm to the left of the horizontal center line of the optical window, and the single-photon avalanche diode is located symmetrically to the right of the center line. Of course, the laser emitter and the single-photon avalanche diode can also be set in other positions, such as the lower edge of the smoke hood, facing directly in front of the smoke hood. This application embodiment does not limit the position of the laser emitter and the single-photon avalanche diode.

[0040] When the range hood enters the gesture recognition control mode, the controller triggers the laser emitter to emit a laser signal in the target direction. The target direction refers to the approximate area in front of the range hood where the user's hands are likely to appear while cooking in front of the stove, such as within a certain space below the range hood's smoke collection hood with the stove as the center, thus facilitating the recognition of the user's gestures in the target direction.

[0041] When the laser signal propagates to the user's hand, it is reflected, forming a reflected light signal. When the reflected light signal reaches the photosensitive area of ​​the single-photon avalanche diode, the photons are absorbed and electron-hole pairs are generated. Under the influence of a strong reverse bias electric field, these initially generated electron-hole pairs undergo an avalanche multiplication effect, thereby forming a detectable electrical signal.

[0042] Step 120: Generate a depth map including the user's hand based on the time difference between the laser signal and the reflected light signal.

[0043] In this embodiment, when the controller of the smoke machine issues a command to control the laser emitter to emit a laser signal, a high-precision timer, such as a time-to-digital converter or a high-frequency counter, can be started simultaneously. The laser signal propagates towards the target direction. When the laser signal shines on the user's hand, part of the light signal is reflected back by the surface of the hand, forming a reflected light signal. When the single-photon avalanche diode receives the reflected light signal, it immediately generates an electrical signal, which triggers the timer to stop counting. The time difference between the laser signal and the reflected light signal is the time interval between the timer starting (corresponding to the laser signal emission time) and the timer stopping (corresponding to the reflected light signal being received).

[0044] Based on the time difference between the laser signal and the reflected light signal, the one-way distance of the laser signal propagation, which is the distance from the user's hand to the single-photon avalanche diode, can be calculated. The calculation process is as follows:

[0045] Where d represents the one-way distance of the laser signal propagation, and c represents the speed of light. This represents the time difference between the laser signal and the reflected light signal.

[0046] In the embodiments of this application, the positions of the single-photon avalanche diode and the laser emitter are usually fixed. Therefore, the positions of the single-photon avalanche diode and the laser emitter can be used as the origin to establish a coordinate system. The one-way distance of the laser signal propagation can be mapped to the coordinate system as a pixel. The pixels in one frame period constitute a depth map including the user's hand.

[0047] Step 130: Identify user gestures based on the depth map.

[0048] In this application, user gestures are actions or postures made by the user with their hands within the sensory area of ​​the range hood, which have specific meanings and patterns, such as waving, hovering, clicking, swiping, drawing circles, etc.

[0049] Since kitchen walls and cabinets can also reflect light, the depth map will include pixels corresponding to the user's hand, as well as pixels corresponding to walls and cabinets. Therefore, it is necessary to segment the area containing the user's hand from the area containing walls and cabinets in the depth map, and analyze the area containing the user's hand to identify the user's gesture.

[0050] In some embodiments, a distance threshold range can be set to identify regions in the depth map that are close to the single-photon avalanche diode and conform to hand size characteristics as the region where the user's hand is located. Alternatively, region growing or deep learning models (such as semantic segmentation models based on convolutional neural networks) can be used to extract the region where the user's hand is located from the depth map.

[0051] After extracting the area containing the user's hand, features representing the gesture type can be further extracted, such as the center coordinates, outline, area, aspect ratio, number and posture of fingers (extended, bent), and palm orientation. For example, a "hover" gesture is characterized by the center point of the hand area remaining essentially unchanged over a period of time, and the outline shape being stable. A "click" gesture is characterized by a rapid contraction and expansion of the hand outline within a short period of time.

[0052] The extracted features can be input into a classification model, which then determines which predefined user gesture the gesture belongs to. The classification model can employ classic machine learning algorithms such as Support Vector Machines, Decision Trees, and Random Forests, or it can use deep learning algorithms. Alternatively, the extracted features can be compared with the features corresponding to predefined user gestures to calculate similarity and thus determine the type of user gesture.

[0053] Step 140: Generate control commands corresponding to user gestures to control the operating status of the range hood.

[0054] Control commands can include power on / off, gear adjustment, lighting switch, delayed power off, etc.

[0055] In this embodiment, the range hood controller can store a gesture command mapping table, which defines the correspondence between each user gesture and its corresponding control command. The mapping relationship can be one-to-one, one-to-many, or many-to-one. For example, a user gesture of "waving from left to right" can uniquely correspond to the command "increase fan speed by one level"; a user gesture of "hovering for two seconds" can correspond to either a "power on" or "power off" command, depending on the current operating state of the range hood. For instance, when the range hood is in standby mode, a user gesture of "hovering for two seconds" corresponds to the power on control command, while when the range hood is running, a user gesture of "hovering for two seconds" corresponds to the power off control command.

[0056] After determining the user's gesture, the corresponding control command can be matched through the gesture command mapping table, and the operating status of the range hood can be controlled according to the control command.

[0057] According to the gesture recognition-based range hood control method of this application, a laser emitter is controlled to emit a laser signal in the target direction of the range hood, and a single-photon avalanche diode is used to receive the reflected light signal formed by the laser signal after reflection from the user's hand. A depth map including the user's hand is generated based on the time difference between the laser signal and the reflected light signal. The user's gesture is recognized based on the depth map. A control command corresponding to the user's gesture is generated to control the operating state of the range hood. This embodiment of the application configures a single-photon avalanche diode and a laser emitter on the range hood. The laser emitter emits a laser signal in the target direction, and the single-photon avalanche diode captures the reflected light signal formed by reflection from the user's hand. Based on the time difference between the laser signal and the reflected light signal, a depth map reflecting the spatial position and morphological characteristics of the user's hand can be generated. Thus, the user's gesture can be recognized based on the depth map, and a control command corresponding to the recognition result can be generated to regulate the operating state of the range hood. This achieves non-contact control of the range hood, reduces the problem of oil contamination of the range hood components by the user's hands during cooking, and improves the convenience of operating the range hood.

[0058] In some embodiments, the single-photon avalanche diode is an array-type single-photon avalanche diode, which corresponds to multiple pixels, and each pixel corresponds to an independent single-photon avalanche diode; the method further includes: When the laser emitter emits a laser signal toward the target direction of the smoke machine, the emission time is recorded by the pixel timing circuit; When a laser signal is received by a single-photon avalanche diode and reflected by the user's hand, the receiving time is recorded by the timing circuit of the corresponding pixel of the single-photon avalanche diode. The time difference between the laser signal and the reflected light signal is calculated based on the transmission and reception times.

[0059] In this embodiment, to improve the accuracy of reflected light signal acquisition, the single-photon avalanche diode is an array-type single-photon avalanche diode. Specifically, the array-type single-photon avalanche diode may include Q×R independent pixels arranged in rows and columns, that is, Q rows and R columns of independent pixels arranged in a two-dimensional row and column configuration, where Q and R are positive integers greater than or equal to 1. The specific values ​​can be set according to the comprehensive consideration of factors such as spatial resolution, field of view, and cost in practical applications. For example, in one example, Q can be 32 and R can be 32, forming a 32×32 single-photon avalanche diode.

[0060] Each pixel may include a single-photon avalanche diode sensing area for converting incident photons into electrical signals, and may also include timing circuitry. The timing circuitry may include, but is not limited to, a high-precision time-to-digital converter, a high-frequency oscillator-driven counter, etc. Through the timing circuitry, each pixel in the array of single-photon avalanche diodes can record the corresponding laser signal emission time and reflected light signal reception time.

[0061] When the laser emitter emits a laser signal toward the target direction of the smoke machine according to the emission command, the timing circuit corresponding to each pixel unit in the array of single-photon avalanche diodes will be synchronously activated by a global synchronization signal or the trigger signal of the laser emitter, and the emission time t1 at this time will be recorded.

[0062] When the laser signal propagates to the user's hand and is reflected, the resulting reflected light signal returns towards the smoke machine. When photons in the reflected light signal reach the array of single-photon avalanche diodes and are absorbed by the sensing area of ​​one or more pixels of the single-photon avalanche diode, an electrical pulse signal is generated. When the timing circuit detects the electrical pulse signal, it triggers a timing stop and records the stop time as the reception time. Then, the transmission time and reception time are subtracted to obtain the time difference between the laser signal and the reflected light signal.

[0063] It should be noted that since the user's hand is a three-dimensional object, the distance from each point on the surface of the user's hand to the array of single-photon avalanche diodes is different. Therefore, the optical path length of the laser signal emitted from the laser emitter after being reflected at different positions on the hand and returning to different pixels in the array is also different. This results in slight differences in the reception time recorded by different pixels. The depth information of each point on the hand can be calculated through these differences.

[0064] In this way, each pixel of the arrayed single-photon avalanche diode has a corresponding time difference, and the collection of time differences of all pixels constitutes a time difference image.

[0065] In this embodiment, by configuring an array of single-photon avalanche diodes and a laser emitter on the tobacco machine, with each array of single-photon avalanche diodes having multiple pixels and each pixel having an independent single-photon avalanche diode and timing circuit, the timing circuit of each pixel can synchronously record the laser emission time when the laser emitter emits a laser signal toward the target direction of the tobacco machine. When the array of single-photon avalanche diodes receives the reflected light signal formed by the user's hand, the timing circuit of the corresponding pixel can capture the received time of the reflected light, thereby improving the accuracy of time difference calculation.

[0066] In some embodiments, generating a depth map including the user's hand based on the time difference between the laser signal and the reflected light signal includes: Point cloud data representing the user's hand position in three-dimensional space is calculated based on the time difference between the laser signal and the reflected light signal, resulting in a depth map including the user's hand.

[0067] In this embodiment, the distance d from each position of the user's hand to the single-photon avalanche diode can be calculated based on the time difference between the laser signal and the reflected light signal. The positions of the single-photon avalanche diode and the laser emitter are usually fixed. Therefore, a coordinate system can be established using the positions of the single-photon avalanche diode and the laser emitter as the origin. The one-way distance of the laser signal propagation is mapped to the coordinate system as a pixel. The pixels within one frame period constitute a depth map containing the user's hand.

[0068] In this embodiment, point cloud data representing the three-dimensional spatial position of the user's hand is calculated by the time difference between the laser signal and the reflected light signal. The point cloud data can represent the three-dimensional shape, spatial coordinates and contour details of the user's hand in three-dimensional space, so that the generated depth map can accurately reflect the relative positional relationship and spatial distribution characteristics of various parts of the hand.

[0069] In some embodiments, point cloud data characterizing the user's hand position in three-dimensional space is calculated based on the time difference between the laser signal and the reflected light signal, including: The distance between the user's hand and the single-photon avalanche diode is calculated based on the time difference between the laser signal and the reflected light signal. Based on the position information and distance of the pixels corresponding to the single-photon avalanche diode, point cloud data representing the position of the user's hand in three-dimensional space is generated.

[0070] In this embodiment, after obtaining the time difference corresponding to each pixel in the array-type single-photon avalanche diode, the radial distance d from the point on the user's hand surface corresponding to each pixel to the single-photon avalanche diode can be calculated based on the time difference.

[0071] To construct a complete three-dimensional coordinate system, the inherent positional information of each pixel within the array of single-photon avalanche diodes can be incorporated. Specifically, each pixel in the array has predefined two-dimensional coordinates (u, v), typically with the top-left corner of the array as the origin, the u-axis pointing horizontally to the right, and the v-axis pointing vertically downwards. These coordinates can be the pixel's row and column indices, for example, the pixel in the i-th row and j-th column, represented as (i, j), or they can be calibrated physical coordinates, for example, in millimeters, representing the horizontal and vertical offset of the pixel center relative to the array origin.

[0072] Using the two-dimensional coordinates (u, v) of a pixel and the corresponding distance d, and combining the intrinsic parameters of an array of single-photon avalanche diodes, a geometric transformation algorithm is used to calculate the coordinates (X, Y, Z) of the hand surface point corresponding to each pixel in the three-dimensional world coordinate system. The intrinsic parameters may include, but are not limited to: the physical size of the pixel (such as pixel pitch), effective focal length, principal point coordinates (the intersection of the optical axis and the sensor plane), etc., and can be obtained in advance through a pre-calibration process.

[0073] In one example, point cloud data can be calculated based on the following simplified geometric transformation model: assuming that the origin of the 3D world coordinate system coincides with the optical center of the arrayed single-photon avalanche diode, the Z-axis is perpendicular to the array plane and points in the user direction (i.e., the depth direction), the X-axis is horizontal to the right, and the Y-axis is vertically downward.

[0074] X = ((u - c_u) × s_x ) × d / f Y = ( (v - c_v) × s_y ) × d / f Z = d Where: (c_u, c_v) represent the principal point coordinates; s_x and s_y represent the physical dimensions of the pixel in the u-axis and v-axis directions, respectively; f represents the effective focal length; and d represents the radial distance from the point on the user's hand surface to the single-photon avalanche diode.

[0075] Through the above transformation, the pixels in the array of single-photon avalanche diodes can be converted into points (X, Y, Z) in three-dimensional space. The set of all three-dimensional points constitutes point cloud data that can characterize the position, contour and shape of the user's hand in three-dimensional space.

[0076] The 3D point cloud data can be orthogonally projected onto an X–Y reference plane parallel to the front panel of the range hood, and voxelized using a grid of preset size. The mean Z-axis value of each point within the grid is calculated to generate an 8-bit grayscale depth map. The grayscale values ​​in the depth map correspond to the depth range, and the area where the user's hand is located can be highlighted and raised in the image.

[0077] In this embodiment, point cloud data is generated by calculating the distance between each position of the user's hand and the single-photon avalanche diode, and then combining the pixel position information. This provides radial depth and spatial coordinate references for the generation of point cloud data, realizing the mapping of three-dimensional coordinates of each position of the hand, so that the generated point cloud data can accurately reproduce the three-dimensional structure, contour details and spatial posture of the hand.

[0078] In some embodiments, recognizing user gestures based on depth maps includes: Obtain depth maps at multiple consecutive time points; Identify the user's hand movement trajectory based on depth maps from multiple consecutive time points; The user's gestures are determined based on the movement trajectory.

[0079] In this embodiment, user gestures can include static gestures and dynamic gestures. Dynamic gestures are essentially the movement of the hand over a period of time. Therefore, it is difficult to capture complete gesture information using only a single frame depth map; it is necessary to obtain depth maps from multiple consecutive moments to obtain a depth map time series.

[0080] Specifically, a single-photon avalanche diode can continuously acquire reflected light signals at a preset frame rate and obtain a depth map based on the reflected light signals. For example, the frame rate can be set to 30 frames per second, or other values. Each frame of acquired reflected light signals can correspond to a depth map, and the continuously acquired depth maps in chronological order constitute a depth image time series.

[0081] After obtaining the depth image time series, the positional changes of the user's hand can be extracted and tracked from the depth image time series, thereby identifying the motion trajectory.

[0082] Specifically, for each frame of the depth map in a depth image time series, the hand region can be located using a hand detection algorithm. For example, the depth map provides distance information, and the hand region appears as a connected region with a significant depth difference from the background. Analyzing the distance information in the depth map allows for the location of the hand region. After detecting the hand region, the coordinates of its center point can be extracted as the representative position of the hand in that frame's depth map. Tracking algorithms, such as Kalman filtering, particle filtering, or deep learning-based tracking models, can be used to associate the hand region detected in the current frame's depth map with the hand region tracked in the previous frame's depth map, thereby tracking the same hand in consecutive frame depth maps.

[0083] In addition to tracking the overall center point of the hand, feature points such as fingertips and knuckles can be extracted within the hand region. By matching these feature points across consecutive depth maps, the movement and posture changes of the hand can be described more accurately. By recording the coordinates of the hand's center point or other feature points in multiple consecutive depth maps and connecting these coordinates in chronological order, the user's hand's motion trajectory in space can be generated. This trajectory is a temporal set of coordinate points that reflects information such as the path, direction, and speed of the hand's movement.

[0084] After successfully extracting the user's hand movement trajectory, the trajectory can be compared with a pre-set gesture template to determine the user's gesture.

[0085] Specifically, trajectory features can be extracted from the motion trajectory, such as the shape of the trajectory, the direction of motion, the distance of motion, the changes in speed and acceleration, and the positional relationship between the starting and ending points of the trajectory.

[0086] The gesture template stores template trajectories and corresponding feature vectors for various user gestures. The extracted trajectory features can be compared with the feature vectors of various user gestures in the gesture template to calculate similarity; for example, Euclidean distance can be used. Based on the similarity calculation results, the user gesture corresponding to the feature vector with the highest similarity to the trajectory features can be identified as the user gesture corresponding to that motion trajectory.

[0087] In this embodiment, by acquiring depth maps at multiple consecutive moments, it is possible to capture the spatial position and shape changes of the user's hand at different points in time, reconstruct the dynamic change process of the hand from a time dimension, thereby improving the accuracy of recognizing the user's hand movement trajectory and thus improving the accuracy of the user's gestures.

[0088] In some embodiments, the user gesture is identified based on the depth map to obtain the user gesture, including: Obtain depth maps at multiple consecutive time points; The depth maps at multiple consecutive time points are input into the gesture recognition model to obtain the user gestures output by the gesture recognition model.

[0089] In this embodiment, the gesture recognition model is pre-built based on deep learning algorithms. For example, convolutional neural networks, recurrent neural networks, long short-term memory networks (LSTM) can be used to build the gesture recognition model. Then, the gesture recognition model is trained so that the trained gesture recognition model can complete the task of recognizing user gestures.

[0090] During the training phase, a large number of depth image time series samples labeled with user gesture categories can be used for training. For example, different user gesture actions can be performed, and depth maps at multiple consecutive time points can be generated based on the time difference between laser signals and reflected light signals to form depth image time series samples. For each depth image time series, a label representing the user gesture category can be established.

[0091] During training, the loss value between the predicted value of the user's gesture output and the label is calculated. Based on the loss value, the internal parameters of the gesture recognition model are continuously adjusted through the backpropagation algorithm to minimize the error between its predicted output and the label. After sufficient training, the gesture recognition model has the ability to analyze and recognize user gestures from new and unseen depth image time series.

[0092] In this embodiment, depth maps from multiple consecutive time points can be input into a trained gesture recognition model. The gesture recognition model processes the time series of the depth images layer by layer and outputs the user's gesture. For example, the gesture recognition model can extract spatial features through structures such as convolutional layers to capture local features and shape information in the depth map; then, it can capture the temporal dynamic information in the time series of the depth map through structures such as recurrent layers to understand the temporal features of hand movements. Finally, the output layer of the gesture recognition model outputs a probability distribution, where each probability value corresponds to a user gesture, and the user gesture with the highest probability can be selected as the recognition result of the gesture recognition model.

[0093] In this embodiment, by acquiring depth maps at multiple consecutive moments, the spatial position and shape changes of the user's hand at different points in time can be captured, and the dynamic change process of the hand can be reconstructed from the time dimension. The continuous depth maps are input into the gesture recognition model, and the gesture recognition model can extract high-order features from the depth maps through deep learning algorithms, thereby accurately recognizing the user's gestures.

[0094] The gesture recognition-based range hood control method provided in this application can be executed by a gesture recognition-based range hood control device. This application uses the example of a gesture recognition-based range hood control device executing the gesture recognition-based range hood control method to illustrate the gesture recognition-based range hood control device provided in this application.

[0095] This application also provides a gesture recognition-based control device for a range hood, wherein the range hood is equipped with a single-photon avalanche diode and a laser emitter.

[0096] like Figure 2 As shown, the gesture recognition-based range hood control device includes: The transmitting and receiving module 210 is used to control the laser transmitter to transmit laser signals toward the target direction of the smoke machine, and to receive the reflected light signal formed by the laser signal after being reflected by the user's hand through a single-photon avalanche diode. The generation module 220 is used to generate a depth map including the user's hand based on the time difference between the laser signal and the reflected light signal; Recognition module 230 is used to recognize user gestures based on depth maps; The control module 240 is used to generate control commands corresponding to user gestures to control the operating status of the range hood.

[0097] According to the gesture recognition-based range hood control device of this application, a laser emitter is controlled to emit a laser signal in the target direction of the range hood, and a single-photon avalanche diode is used to receive the reflected light signal formed by the laser signal after reflection from the user's hand. A depth map including the user's hand is generated based on the time difference between the laser signal and the reflected light signal. The user's gesture is recognized based on the depth map. A control command corresponding to the user's gesture is generated to control the operating state of the range hood. The embodiment of this application configures a single-photon avalanche diode and a laser emitter on the range hood. The laser emitter emits a laser signal in the target direction, and the single-photon avalanche diode captures the reflected light signal formed by reflection from the user's hand. Based on the time difference between the laser signal and the reflected light signal, a depth map reflecting the spatial position and shape characteristics of the user's hand can be generated. Thus, the user's gesture can be recognized based on the depth map, and a control command corresponding to the recognition result can be generated to regulate the operating state of the range hood. This achieves non-contact control of the range hood, reduces the problem of oil and grease from the user's hands contaminating the range hood components during cooking, and improves the convenience of operating the range hood.

[0098] In some embodiments, the transmitting and receiving module 210 is further configured to: When the laser emitter emits a laser signal toward the target direction of the smoke machine, the emission time is recorded by the pixel timing circuit; When a laser signal is received by a single-photon avalanche diode and reflected by the user's hand, the receiving time is recorded by the timing circuit of the corresponding pixel of the single-photon avalanche diode. The time difference between the laser signal and the reflected light signal is calculated based on the transmission and reception times.

[0099] In some embodiments, the generation module 220 is further configured to: Point cloud data representing the user's hand position in three-dimensional space is calculated based on the time difference between the laser signal and the reflected light signal, resulting in a depth map including the user's hand.

[0100] In some embodiments, the generation module 220 is further configured to: The distance between the user's hand and the single-photon avalanche diode is calculated based on the time difference between the laser signal and the reflected light signal. Based on the position information and distance of the pixels corresponding to the single-photon avalanche diode, point cloud data representing the position of the user's hand in three-dimensional space is generated.

[0101] In some embodiments, the identification module 230 is further configured to: Obtain depth maps at multiple consecutive time points; Identify the user's hand movement trajectory based on depth maps from multiple consecutive time points; The user's gestures are determined based on the movement trajectory.

[0102] In some embodiments, the identification module 230 is further configured to: Obtain depth maps at multiple consecutive time points; The depth maps at multiple consecutive time points are input into the gesture recognition model to obtain the user gestures output by the gesture recognition model.

[0103] The gesture recognition-based range hood control device in this application embodiment can be an electronic device or a component within an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, PDA, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc., and can also be a server, a controller in the range hood, or a functional module, etc. This application embodiment does not specifically limit the specific implementation.

[0104] The gesture recognition-based range hood control device in this application embodiment can be a device with an operating system. This operating system can be a Microsoft (Windows) operating system, an Android operating system, an embedded system, or other possible operating systems; this application embodiment does not specifically limit it.

[0105] In some embodiments, this application provides a smoke hood, including a single-photon avalanche diode, a laser emitter, and a controller; The controller is used to execute the various processes of the above-described embodiment of the gesture recognition-based smoke hood control method, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0106] In some embodiments, such as Figure 3 As shown, this application embodiment also provides an electronic device 300, including a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301. When the program is executed by the processor 301, it implements the various processes of the above-described embodiment of the gesture recognition-based smoke machine control method and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0107] It should be noted that the electronic devices in the embodiments of this application include the aforementioned mobile electronic devices and non-mobile electronic devices.

[0108] This application also provides a non-transitory computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described gesture recognition-based smoke machine control method embodiment and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0109] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0110] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described gesture recognition-based smoke machine control method.

[0111] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0112] This application also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described embodiments of the gesture recognition-based smoke machine control method, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0113] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0114] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0115] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0116] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0117] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0118] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for controlling a range hood based on gesture recognition, the method comprising: The smoke machine is equipped with a single-photon avalanche diode and a laser emitter; the method includes: The laser emitter is controlled to emit a laser signal toward the target direction of the smoking machine, and the reflected light signal formed by the laser signal after being reflected by the user's hand is received by the single-photon avalanche diode. A depth map including the user's hand is generated based on the time difference between the laser signal and the reflected light signal; User gestures are identified based on the depth map; The system generates control commands corresponding to the user's gestures to control the operating status of the range hood.

2. The method according to claim 1, characterized in that, The single-photon avalanche diode is an array-type single-photon avalanche diode, which corresponds to multiple pixels, and each pixel corresponds to an independent single-photon avalanche diode; the method further includes: When the laser emitter emits a laser signal toward the target direction of the smoke machine, the emission time is recorded by the timing circuit of the pixel; When the single-photon avalanche diode receives the reflected light signal formed by the laser signal reflected by the user's hand, the receiving time is recorded by the timing circuit of the pixel corresponding to the single-photon avalanche diode; The time difference between the laser signal and the reflected light signal is calculated based on the transmission time and the reception time.

3. The method according to claim 1, characterized in that, The step of generating a depth map including the user's hand based on the time difference between the laser signal and the reflected light signal includes: The point cloud data representing the three-dimensional spatial position of the user's hand is calculated based on the time difference between the laser signal and the reflected light signal, resulting in a depth map including the user's hand.

4. The method according to claim 3, characterized in that, The calculation of point cloud data representing the three-dimensional spatial position of the user's hand based on the time difference between the laser signal and the reflected light signal includes: The distance between each position of the user's hand and the single-photon avalanche diode is calculated based on the time difference between the laser signal and the reflected light signal. Based on the position information of the pixel corresponding to the single-photon avalanche diode and the distance, point cloud data representing the position of the user's hand in three-dimensional space is generated.

5. The method according to claim 3, characterized in that, The step of recognizing user gestures based on the depth map includes: Obtain depth maps at multiple consecutive time points; The user's hand movement trajectory is identified based on depth maps at multiple consecutive time points; The user's gesture is determined based on the motion trajectory.

6. The method according to claim 1, characterized in that, The step of identifying the user gesture based on the depth map to obtain the user gesture includes: Obtain depth maps at multiple consecutive time points; The depth map at multiple consecutive time points is input into the gesture recognition model to obtain the user gesture output by the gesture recognition model.

7. The method according to claim 1, characterized in that, The user gestures include at least one of waving, hovering, clicking, swiping, and drawing a circle; the control commands include at least one of power on, power off, gear adjustment, lighting switch, and delayed power off.

8. A smoke machine control device based on gesture recognition, characterized in that, The smoke machine is equipped with a single-photon avalanche diode and a laser emitter; the device includes: The transmitting and receiving module is used to control the laser transmitter to emit a laser signal toward the target direction of the smoke machine, and to receive the reflected light signal formed by the laser signal after being reflected by the user's hand through the single-photon avalanche diode. The generation module is used to generate a depth map including the user's hand based on the time difference between the laser signal and the reflected light signal; The recognition module is used to recognize user gestures based on the depth map; The control module is used to generate control commands corresponding to the user's gestures in order to control the operating status of the smoke machine.

9. A range hood, characterized in that, Includes single-photon avalanche diodes, laser emitters, and controllers; The controller is configured to perform the method as described in any one of claims 1-7.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-7.

Images