Clothes airing machine control method and system based on gestures, medium and equipment

Abstract

The invention discloses a gesture-based clothes airing machine control method and system, a medium and equipment, the method is applied to a clothes airing machine provided with an image acquisition unit, and the method comprises the following steps: acquiring image information acquired by the image acquisition unit; detecting whether face information exists according to the image information; if the face information exists in the image information, extracting a target gesture from the image information; identifying the extracted target gesture, and determining a gesture action; and controlling the clothes airing machine to execute a corresponding function according to the determined gesture action. According to the method, the clothes airing machine is controlled through the gesture action, specifically, gesture recognition is started under the condition that the face information is detected, the situation that a user who enters the airing induction area and does not have control consciousness triggers the clothes airing machine by mistake due to the coincident gesture action can be avoided, the recognition accuracy of the clothes airing machine is effectively improved, and the user experience is improved. And the situation of false triggering is avoided.

Description

Technical Field

[0001] The present application relates to the technical field of clothes drying control, and in particular to a method, system, medium and device for controlling a clothes drying machine based on gestures. Background Art

[0002] Smart clothes drying machines are becoming more and more popular because they can be automatically raised and lowered and have integrated functions such as lighting, drying, and sterilization. They have become an indispensable part of smart home products.

[0003] At present, clothes drying racks on the market are basically controlled by handheld wireless remote control, APP, and voice. The wireless remote control communicates with the host through the RF wireless module in a one-to-one mode to control the corresponding functions; the APP is based on the network platform through the mobile phone and opens the corresponding APP, and the corresponding function can be started and stopped by clicking the screen; voice control, the function of voice interaction with the clothes drying rack, specifying the wake-up word to shout to the clothes drying rack, and speaking the specified words to control the function of the clothes drying rack. Among these control methods, the wireless remote control is placed randomly, and there is a situation where the remote control needs to be found frequently; and the APP control requires the clothes drying rack to be connected to the Internet, and requires a smart terminal such as a mobile phone to operate and control; the voice recognition control method has the problems of recognition delay and low recognition rate, especially for people who do not speak standard Mandarin, even if the voice is repeated, it cannot be recognized, and the user experience is poor. In addition, due to the wide range of voice information collection, the voice recognition control method is easy to misidentify the control information from the information that the user usually speaks. Summary of the invention

[0004] The first purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and to provide a gesture-based clothes drying machine control method. The method of the present invention enables the clothes drying machine to control corresponding functions based on gesture actions; it can effectively improve the accuracy of clothes drying machine recognition and avoid false triggering.

[0005] A second object of the present invention is to provide a clothes drying machine control system.

[0006] A third object of the present invention is to provide a storage medium.

[0007] A fourth object of the present invention is to provide an electronic device.

[0008] The first object of the present invention is achieved by the following technical solution: A gesture-based clothes drying machine control method, applied to a clothes drying machine equipped with an image acquisition unit, comprising the steps of:

[0009] Acquiring image information acquired by an image acquisition unit;

[0010] For image information, detect whether there is face information;

[0011] If there is face information in the image information, extract the target gesture from the image information;

[0012] Identify the extracted target gesture and determine the gesture action;

[0013] The clothes drying machine is controlled to execute corresponding functions according to the determined gesture actions.

[0014] Preferably, for the image information collected by the image acquisition unit, when multiple gestures are extracted, the distance between each gesture and the face is calculated based on each gesture area and face avatar area in the image information, and the gesture closest to the face is selected as the target gesture.

[0015] Furthermore, the distance between each gesture and the face is calculated in the following way:

[0016]

[0017] Among them, (x i ,y i ) is the coordinate of the target point i in the face image in the image, (x s ,y s ) is the coordinate of the gesture target point s in the image; D is is the distance between the gesture target point s and the face avatar target point i in the image.

[0018] Preferably, the method further comprises the following steps:

[0019] Setting a dedicated gesture action; wherein the dedicated gesture action is used to control the clothes drying machine to execute the function of lowering the clothes drying rod;

[0020] For special gesture actions, configure human height information;

[0021] When the gesture action in the image information is recognized as a special gesture action, the configured human body height information is obtained, and the clothes drying rod of the clothes drying machine is controlled to descend to a corresponding stroke according to the human body height information.

[0022] Preferably, the method further comprises the following steps of detecting the height of a human body:

[0023] For the image information collected by the image acquisition unit, extract the face image from the image information, and perform depth map calculation to obtain a depth map calculation result;

[0024] Calculate the height h of the human body based on the depth map calculation result:

[0025] h=M-(Z i cosα+Y i sinα);

[0026] in:

[0027] M is the installation height of the image acquisition unit on the clothes drying machine, α is the inclination angle of the image acquisition unit installation, which corresponds to the angle between the central axis of the image acquisition unit and the plumb line;

[0028] Z i and Y i is the depth map calculation result, Z i is the depth coordinate value of the target point i of the face portrait, Y i is the height coordinate value of target point i of the face portrait;

[0029] When the gesture action is to control the clothes drying machine to execute the function of lowering the clothes drying rod, the clothes drying machine clothes drying rod is controlled to lower the corresponding stroke according to the human body height detected in the above-mentioned human body height detection step.

[0030] Furthermore, the image acquisition unit is a binocular camera installed on the clothes drying machine at a set angle, including a first camera and a second camera; a left image and a right image acquired by the first camera and the second camera are acquired respectively;

[0031] Detecting facial information, determining gestures, and obtaining depth map calculation results by any of the following methods:

[0032] The first one:

[0033] The left image and the right image are cropped respectively;

[0034] For the cropped left image, normalize it and detect the face information;

[0035] When face information is detected in the left image, the target gesture is extracted from the normalized left image;

[0036] Classify and identify the extracted target gestures to determine the gesture actions;

[0037] For the cropped right image, normalize it and detect the face information;

[0038] When face information is detected in the right image, extract the left face image and the right face image respectively for the left image and the right image;

[0039] For the face portrait in the left image and the face portrait in the right image, depth map calculation is performed to obtain depth map calculation results;

[0040] Second type:

[0041] The left image and the right image are cropped respectively;

[0042] For the cropped right image, normalize it and detect the face information;

[0043] When face information is detected in the right image, target gesture features are extracted from the normalized right image;

[0044] Classify and identify the extracted target gesture features to determine the gesture action;

[0045] For the cropped left image, normalize it and detect the face information;

[0046] When face information is detected in the left image, extracting a left face avatar and a right face avatar from the left image and the right image respectively;

[0047] A depth map calculation is performed for the left-hand face portrait and the right-hand face portrait to obtain a depth map calculation result.

[0048] Furthermore, in the human height detection step:

[0049] When a human face image is detected in the image information, the human face image is used as the target human face image, and a depth map calculation is performed to obtain a depth map calculation result;

[0050] When multiple face images are extracted from the image information, the target face image is determined by any of the following methods, and a depth map is calculated for the target face image to obtain a depth map calculation result:

[0051] The first one:

[0052] Obtaining the coordinate value of each face avatar target point in the image information at depth, and determining the depth information of each face avatar through the coordinate value of each face avatar target point at depth; then comparing the depth information of each face avatar, and determining the target face avatar based on the face avatar with the smallest depth distance;

[0053] Second type:

[0054] Obtain the coordinate value of the height of each face avatar target point in the image information, and determine the height information of each face avatar through the coordinate value of the height of each face avatar target point; then compare the height information of each face avatar, and determine the target face avatar based on the face avatar with the smallest height distance;

[0055] The third type:

[0056] For each face avatar extracted from the image information, a depth map is calculated respectively, and then the corresponding human height is detected through the human height detection step; the human heights corresponding to each face avatar are compared, and the target face avatar is determined based on the face avatar with the shortest human height.

[0057] Furthermore, when the image acquisition unit is a binocular camera, in the human height detection step: obtain the left image and the right image acquired by the binocular camera; for the left image and the right image, determine whether the face portraits extracted therein meet the following conditions: in the pixel coordinates (x, y) of the top corner of the face frame, |y| is less than a set value; if so, determine the extracted face portrait as the target face image, otherwise determine that there is no target face image in the image;

[0058] In the human body height detection step, a depth map is calculated for the target face portrait to obtain a depth map calculation result.

[0059] Furthermore, the human body height detection step also includes a human body height compensation step, which is achieved by any of the following:

[0060] For the face image area extracted from the image information, determine the pixel point j at the highest position of the face image in the image, then calculate the distance h" between the face image target point i and the pixel point j in the image, and add the h value to h" as the final human height detection result;

[0061] Set h" as a fixed compensation value, and add the h value to h" as the final human height detection result.

[0062] The second object of the present invention is achieved by the following method: A clothes drying machine control system includes a controller, an image processor and an image acquisition unit;

[0063] The image acquisition unit is used to acquire image information;

[0064] The image processor is connected to the image acquisition unit and is used to detect face information from the image information acquired by the image acquisition unit, extract target gesture features from the image information when it is determined that face information is detected in the image information, and identify the extracted target gesture features to determine the gesture action;

[0065] The controller is connected to the image processor and is used to obtain the gesture action determined by the image processor, and control the clothes drying machine to perform corresponding functions according to the determined gesture action.

[0066] Preferably, it also includes a human body sensing unit;

[0067] The human body sensing unit is connected to the controller and is used to detect whether a person enters the drying sensing area. When a person enters, the controller starts the image acquisition unit to collect image information.

[0068] The third object of the present invention is achieved by the following method: a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the gesture-based clothes drying machine control method described in the first object of the present invention.

[0069] The fourth purpose of the present invention is achieved in the following manner: an electronic device comprises a processor and a memory for storing programs executable by the processor, and when the processor executes the program stored in the memory, the gesture-based clothes drying machine control method described in the first purpose of the present invention is implemented.

[0070] Compared with the prior art, the present invention has the following advantages and effects:

[0071] (1) The gesture-based clothes drying machine control method of the present invention is applied to a clothes drying machine equipped with an image acquisition unit. First, the image information acquired by the image acquisition unit is obtained; for the image information, whether there is facial information is detected. If there is facial information, the target gesture is extracted from the image information; the extracted target gesture is identified and classified to determine the gesture action; and the clothes drying machine is controlled to perform the corresponding function according to the determined gesture action. As can be seen from the above, the method of the present invention controls the clothes drying machine through gesture actions, specifically, gesture recognition is started when facial information is detected. Since the image acquisition device installed on the clothes drying machine is located above the user's head, the user needs to consciously look up to collect facial information. Therefore, it can avoid the situation where a user who enters the drying sensing area without control awareness accidentally triggers the clothes drying machine due to coincidental gesture actions, effectively improving the accuracy of clothes drying machine recognition and avoiding the situation of false triggering.

[0072] (2) In the gesture-based clothes drying machine control method of the present invention, for the image information collected by the image acquisition unit, when multiple gestures are extracted, the distance between each gesture and the face is calculated, and the gesture closest to the face is selected as the target gesture; the above processing can effectively ensure that the recognized gesture comes from a user with a real control purpose, that is, the target person, and thus can further reduce the situation of false triggering.

[0073] (3) In the gesture-based clothes drying machine control method of the present invention, a special gesture action can be set for the user. The special gesture action is a gesture action for controlling the clothes drying machine to execute the function of lowering the clothes drying rod. The corresponding human height information is configured for the special gesture action. When the special gesture action is recognized during the gesture recognition process, the clothes drying rod of the clothes drying machine can be controlled to be lowered to the corresponding height based on the configured human height information, so as to realize adaptive control of the lowering of the clothes drying rod, so that the clothes drying rod is lowered to a height suitable for the user to dry clothes. In addition, after extracting the face information from the image, the present invention can also perform depth map calculation, and directly calculate the human height according to the depth map calculation result. Based on this, for each user whose face avatar enters the drying sensing area and is detected, the present invention can realize height detection for them. Further, when it is detected that the function matched by the gesture action is to lower the clothes drying rod, the clothes drying rod can be controlled to be lowered to a suitable height for each user according to the user's human height, so as to realize adaptive adjustment of the lowering height of the clothes drying rod, so as to facilitate different users to dry and collect clothes.

[0074] (4) In the gesture-based clothes drying machine control method of the present invention, the image acquisition unit installed in the clothes drying machine can use a binocular camera, wherein obtaining the image information acquired by the image acquisition unit refers to respectively acquiring the left image and the right image acquired by the first camera and the second camera, and performing cropping processing on the acquired left image and the right image respectively; then, detecting the face information through one of the images (the left image or the right image), performing gesture recognition in the image based on the detection result, detecting the face information through the other image (the right image or the left image), and after detecting the face image, performing depth map calculation based on the face portraits of the left image and the right image, and finally obtaining the depth calculation result. It can be seen that the present invention combines the left image and the right image of the binocular camera to realize face detection before gesture recognition, gesture recognition, and face detection before depth calculation, which can effectively improve the recognition rate of face and gesture recognition, and thus improve the accuracy of control of the clothes drying machine; because depth calculation is only performed when faces are detected in both the left image and the right image, and then human height detection is performed, this method can ensure that human height detection can only be started when the user is in the shooting range of the first camera and the second camera at the same time, which can further avoid the occurrence of false detection situations and improve the accuracy of the control of the lowering of the clothes drying machine.

[0075] (5) In the gesture-based clothes drying machine control method of the present invention, when performing human height detection, if the image information includes multiple face avatars, the present invention can determine the final target person based on the depth coordinates, height coordinates or the comparison results of the human heights corresponding to each face avatar, so that the target person can be automatically identified when multiple people enter the drying sensing area, thereby realizing adaptive control of the height of the clothes drying machine rod. In addition, if the image acquisition unit is a binocular camera, for the acquired left image and right image, it is necessary to determine whether the extracted face avatars meet the following conditions: in the pixel coordinates (x, y) of the top corner of the face frame, |y| is less than the set value; if so, the face avatars extracted from the left image and the right image are determined as the target face image, otherwise it is determined that there is no target face image in the image. Based on this, the method of the present invention can control the face avatars within a certain range acquired by the binocular camera as the target face avatars, improve the flexibility of face avatar acquisition, and adapt to various installation environments.

[0076] (6) In the gesture-based clothes drying machine control method of the present invention, the human body height detection step also includes a human body height compensation process, specifically, the human body height calculation result can be added to the distance h" between the target point i of the face avatar and the pixel point j at the highest position of the face avatar in the left image or the right image; based on this, the human body height calculated by the method of the present invention can be closer to the actual height, further improving the accuracy of height detection. Alternatively, h" can be set as a fixed compensation value, and the h value and h" are added as the final human body height detection result. Such a compensation method makes the calculation amount smaller and the speed faster in the process of human body height detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 It is a flow chart of the clothes drying machine control method of the present invention.

[0078] Figure 2a , 2b 2c and 2c are schematic diagrams of depth calculation under binocular vision.

[0079] Figure 3 It is a schematic diagram of the human body height detection principle of the present invention.

[0080] Figure 4 It is a detailed flow chart of the clothes drying machine control method in embodiment 1 of the present invention.

[0081] Figure 5 It is a structural block diagram of the control system of the clothes drying machine of the present invention. DETAILED DESCRIPTION

[0082] The embodiments of the present application are described below in conjunction with the drawings in the present application. It should be understood that the implementation methods described below in conjunction with the drawings are exemplary descriptions for explaining the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions of the embodiments of the present application.

[0083] Example 1

[0084] At present, the control of clothes drying machines mainly relies on remote control, APP or voice recognition. Since the remote control is not fixed in position and can be moved, users often need to find the remote control during use. APP control requires smart terminals such as mobile phones with corresponding applications installed. Voice recognition control has low recognition accuracy and is easy to misidentify control information from the user's daily speech information.

[0085] Based on the problems existing in the prior art, this embodiment provides a method for controlling a clothes drying machine based on gestures, which is applied to a clothes drying machine equipped with an image acquisition unit.

[0086] The clothes drying machine referred to in this embodiment generally includes a host and a drying rod. The host is generally installed on the ceiling, and the drying rod is set below the host through a wire rope or other lifting parts. The host is provided with a controller (such as MCU), a motor driver, a motor and other devices. The controller is connected to the motor through the motor driver, and the wire rope or the lifting part is controlled by the motor to perform lifting and lowering movements to realize the automatic lifting and lowering of the drying rod. In addition, the clothes drying machine may also be equipped with functional modules such as lighting, drying, and disinfection, and the controller can issue instructions to realize the control of each functional template.

[0087] In this embodiment, the image acquisition unit can be installed on the main unit of the clothes drying machine at a certain angle, specifically, it can be installed on the main unit housing, or embedded in the main unit housing. Generally, a camera can be selected as the image acquisition unit. In this embodiment, in order to more accurately identify gestures and make the detection range wider, the image acquisition unit can use a binocular camera, which includes a first camera and a second camera. The first camera and the second camera simultaneously perform image acquisition work, and respectively acquire the left image and the right image. After the image acquisition unit is installed on the clothes drying machine, an image acquisition area will be formed, that is, objects in the image acquisition area can be captured by the image acquisition unit. In this embodiment, the image acquisition area is defined as a drying sensing area.

[0088] The gesture-based clothes drying machine control method of this embodiment is as follows: Figure 1 As described in, the steps include:

[0089] S101. Acquire image information acquired by an image acquisition unit.

[0090] In this embodiment, when the image acquisition unit uses a binocular camera, image information captured by the first camera and the second camera is obtained respectively, which are defined as a left image and a right image. In this embodiment, the left image and the right image are cropped respectively.

[0091] S102, detecting whether there is face information in the image information; when there is face information in the image information, further extracting the target gesture from the image information, that is, obtaining the corresponding gesture feature.

[0092] In this embodiment, the face information is detected after normalization processing is performed on the cropped left image; or the face information is detected after normalization processing is performed on the cropped right image.

[0093] In this embodiment, for the image information collected by the image acquisition unit (the left image or the right image after the above-mentioned cropping and normalization processing), the gesture features in the image information are extracted to detect the target gesture. If the image information detects multiple gestures when detecting the target gesture, the distance between each gesture and the face avatar is calculated, and the gesture closest to the face avatar is selected as the target gesture; based on this, it can avoid the situation where other gestures are recognized and cause miscontrol. The details are as follows:

[0094] S102-1, obtain the coordinates of the face portrait target point i in the image (left image or right image), and obtain the coordinates of each gesture target point s in the image. The gesture target point refers to the center of gravity, center point or other specified suitable point of the gesture area in the image (such as a point on a finger, back of hand, or palm). The face portrait target point can be the center of gravity of the face portrait area in the image, the center point of the face portrait area (specifically, it can be the center point of the face frame) or other specified suitable point (such as a point on the nose, eye or forehead).

[0095] S102-2, calculate the distance between each gesture and the face avatar using the following formula:

[0096]

[0097] Among them, (x i ,y i ) is the coordinate of the target point i in the face image in the image, (x s ,y s ) is the coordinate of the gesture target point s in the image; D is is the distance between the gesture target point s and the face avatar target point i in the image. The above coordinates are obtained from the left image or the right image captured by the binocular camera, that is, the coordinates of the face avatar target point i and the gesture target point s in the left image or the right image.

[0098] S102-3, comparing the distances between each gesture and the face avatar, and selecting the gesture corresponding to the shortest distance as the target gesture. That is, in this embodiment, the gesture closest to the face avatar in the left image or the right image is selected as the target gesture.

[0099] In this step, if multiple face avatars are detected in the image, the target face avatar can also be determined by the following step S201. When multiple gestures are detected, the distance between each gesture and the target face avatar is calculated in this step.

[0100] S103, identify the extracted target gesture and determine the gesture action. In this embodiment, a gesture recognition model can be trained based on a neural network. Specifically, multiple groups of gesture features can be obtained, and each group of gesture features can be labeled. The label corresponds to a specific gesture action, and a certain number of training samples are obtained. A neural network model (such as a convolutional neural network model) constructed based on the training samples is trained to obtain a gesture recognition model. In this step, the gesture features corresponding to the target gesture extracted from the image information in step S102 are input into the gesture recognition model, and the gesture action is classified and identified by the gesture recognition model. The gesture action may include raising one or more fingers, making a fist, OK gesture, stretching out the palm, etc. The gesture action to be recognized can be selected according to the actual control needs.

[0101] S104: Control the clothes drying machine to execute corresponding functions according to the determined gesture action.

[0102] In this embodiment, before executing step S104, different gesture actions need to be respectively configured with different functions performed by the clothes drying machine. For example, the first gesture action is configured with a function of raising the clothes drying rod, the second gesture action is configured with a function of lowering the clothes drying rod, the third gesture action is configured with a function of turning on the clothes drying machine lighting, the fourth gesture action is configured with a function of turning off the clothes drying machine lighting, the fifth gesture action is configured with a function of turning on the clothes drying machine disinfection, the sixth gesture action is configured with a function of turning off the clothes drying machine disinfection, and so on.

[0103] In the clothes drying machine control method of this embodiment, the following steps may also be included:

[0104] Sa. Setting a dedicated gesture action for the user; wherein the dedicated gesture action is used to control the clothes drying machine to perform a drying rod lowering function;

[0105] Sb. For special gesture actions, the user's height information can be configured through the user APP.

[0106] Sc. When the gesture action in the image information is recognized as a special gesture action, the configured human height information is obtained, and the clothes drying rod of the clothes drying machine is controlled to descend to a corresponding stroke according to the human height information.

[0107] For example, for user A, setting fist as a special gesture action for the user is used to control the clothes drying machine to perform the function of lowering the clothes drying rod. At the same time, other gesture actions such as gesture action X can also be set to control the clothes drying machine to perform the function of lowering the clothes drying rod, but in the method of this embodiment, the human height information of user A is configured for the fist gesture action. When the target gesture is identified as a fist gesture action, the human height information of user A configured as above is obtained, so that the height of the clothes drying rod of the clothes drying machine is controlled according to the human height information of user A, so that the height of the clothes drying rod is suitable for user A to perform the drying operation, and when the target gesture is identified as gesture action X, the clothes drying rod is controlled to be lowered to a fixed height. Further, a user recognition function can also be added, for example, user A is recognized based on face information. After user A is recognized, if the recognized gesture action is a fist, the height of the clothes drying rod of the clothes drying machine is controlled based on the human height information of user A; in this way, it can be avoided that when other users perform a fist gesture action, the height of the clothes drying rod is not controlled according to the human height information of user A.

[0108] In the clothes drying machine control method of this embodiment, a human height detection step may also be included, that is, the human height of the person entering the drying sensing area may be detected; when the gesture action is recognized to control the lowering of the clothes drying machine rod, the height corresponding to the lowering of the rod may be controlled according to the detected human height information. In this embodiment, the specific process of the human height detection step is as follows:

[0109] S201, extracting a face image from the image information collected by the image acquisition unit, and executing step S202 if a face image is extracted, otherwise stopping the human height detection step.

[0110] In this embodiment, the cropped left image or right image in step S101 is obtained; if the left image is used for face recognition and gesture recognition in step S102, the cropped right image is obtained in this step, and the face portrait is extracted after normalization. If the right image is used for face recognition and gesture recognition in step S102, the cropped left image is obtained in this step, and the face portrait is extracted after normalization.

[0111] In this step, after extracting the face image from the normalized image, it is necessary to further determine the target face image. The specific method is as follows:

[0112] When a face image is detected in the image information, the face image is used as the target face image. In the case where the image acquisition unit is a binocular camera, the face image is determined to be the target face image only when the face image is detected in both the left image and the right image.

[0113] When multiple face images are extracted from the image information, the target face image is determined by any of the following methods:

[0114] The first one:

[0115] Obtaining the coordinate value of each face avatar target point in the image information at depth, and determining the depth information of each face avatar through the coordinate value of each face avatar target point at depth; then comparing the depth information of each face avatar, and determining the target face avatar based on the face avatar with the smallest depth distance;

[0116] Second type:

[0117] Obtain the coordinate value of the height of each face avatar target point in the image information, and determine the height information of each face avatar through the coordinate value of the height of each face avatar target point; then compare the height information of each face avatar, and determine the target face avatar based on the face avatar with the smallest height distance;

[0118] The third type:

[0119] For each face image extracted from the image information, a depth map is calculated respectively, and then the corresponding human height is detected through the human height detection step; the human heights corresponding to each face image are compared, and the target face image is determined based on the face image with the shortest human height;

[0120] Furthermore, in this embodiment, the image acquisition unit is a binocular camera. After extracting the face portrait from the left image and the right image, for the left image and the right image, it is determined whether the extracted face portraits therein meet the following conditions: in the pixel coordinates (x, y) of the top corner of the face frame, |y| is less than a set value; if so, the extracted face portrait is determined as the target face image, otherwise it is determined that there is no target face image in the image. Among them, when the face portrait is extracted from the image information, the face portrait will be framed in the form of a box. After the binocular camera acquires the left image and the right image, the face portraits in the left image and the right image are framed in a box. The box is defined as a face frame, and the top corner of the face frame is the upper left corner, upper right corner, lower left corner or lower right corner of the box. Based on the above operations, the method of this embodiment controls the facial portrait within a certain range captured by the binocular camera as the target facial portrait, thereby improving the flexibility of facial portrait acquisition and adapting to various installation environments; the above setting values ​​can be set specifically according to the actual shooting range requirements.

[0121] In this embodiment, the |y| in the coordinates of the top corners of the face frame corresponding to the face portrait extracted from the left image and the right image is compared with the set value. If both satisfy |y| is less than the set value, the face portrait extracted from the left image and the right image is used as the target face portrait, otherwise it is determined that there is no target face portrait in the image, and the subsequent height detection step is not performed. In this embodiment, for the left image or the right image, before the above determination is performed, multiple face portraits may be extracted from it. In this case, one of the face portraits is first determined based on the target face portrait determination method under the above multiple face portraits, and then based on the determined face portrait, combined with the left image and the right image, the above determination is performed to determine whether there is a target face portrait. Based on the above operation of this embodiment, it can also ensure that the subsequent operation is performed only when there are face portraits in both the left image and the right image, and the range of the face portrait is further limited, that is, the face portrait captured within the corresponding shooting range can start the subsequent height detection.

[0122] S202, extract the target face avatar from the left image and the right image to obtain the left face avatar and the right face avatar; perform depth map calculation on the left face avatar and the right face avatar to obtain the depth map calculation result, that is, determine the coordinate value Z of the target point i of the face avatar at depth i and the coordinate value Y at height i .

[0123] In this embodiment, the above-mentioned facial portrait target point can be the center of gravity of the facial portrait area in the image, the center point of the facial portrait area (specifically, it can be the center point of the face frame) or other specified suitable points (for example, points taken at the nose, eyes or forehead); the center of gravity of the facial portrait area refers to the grayscale center of gravity of the facial portrait area.

[0124] In this step, based on the image captured by the binocular camera, the depth coordinate value Z and the height coordinate value Y of each pixel in the face portrait can be calculated by the existing depth map calculation method, which correspond to the two coordinate values ​​in the three-dimensional coordinates (X, Y, Z) of each pixel in the face portrait. Specifically, the height coordinate value Y and the depth coordinate value Z of each pixel in the face portrait can be obtained by depth calculation respectively:

[0125]

[0126]

[0127] Among them, Figures 2a to 2c As shown in the figure, X is the alignment axis, Y is the height, and Z represents the depth. The line connecting the optical centers of the first camera CL and the second camera CR is aligned with the X axis (parallel). The coordinate value of the pixel point of the face portrait corresponding to the X axis in the left image is x L, the coordinate value of the corresponding X-axis in the right image is x R ; The coordinate value of the face pixel point on the left image corresponding to the Y axis is y L , the corresponding Y-axis coordinate value in the right image is y R b is the baseline, which is the distance between the two cameras, f is the focal length of the first camera and the second camera, and the parallax of the left and right cameras is d=|x L -x R |. The formulas for Z and Y in the above depth calculation are existing formulas, and the derivation principles will not be repeated here.

[0128] In this embodiment, if the target point of the above-mentioned face portrait is the center of gravity of the face portrait area, after obtaining the left image and the right image captured by the first camera and the second camera, the left face portrait and the right face portrait are extracted, and then the height coordinate value and the depth coordinate value of each pixel in the face portrait are determined by the above formula in combination with the internal parameters of the binocular camera (including f, b, etc.), and finally the height coordinate value Y of the center of gravity i of the face portrait can be determined based on the grayscale center of gravity method i and depth coordinate value Z i When the target point of the face avatar is the center point of the face avatar, the height coordinate value and depth coordinate value of the center pixel of the face avatar can also be directly determined as the height coordinate value Y of the target point i of the face avatar. i and depth coordinate value Z i .

[0129] S203, calculating the height h of the human body according to the calculation result of the depth map:

[0130] h=M-(Z i cosα+Y i sinα);

[0131] in:

[0132] M is the installation height of the image acquisition unit on the clothes drying machine, α is the inclination angle of the image acquisition unit installation; it corresponds to the central axis of the binocular camera (such as Figure 3 In this embodiment, the inclination angle α of the binocular camera installation can be 10° to 85°, and the inclination angle α refers to the angle between the central axis of the binocular camera and the plumb line, where the plumb line refers to a line perpendicular to the horizontal plane, that is, a line in the height direction of the human body. In this embodiment, the preferred inclination angle is 30° or about 30°. When it is 30°, the drying sensing area can be a rectangular area of ​​2m*1.2m.

[0133] Z i and Y i is the depth map calculation result, Z iis the depth coordinate value of the target point i of the face portrait, Y i It is the coordinate value of the target point i of the face portrait in height.

[0134] In this embodiment, Figure 3 As shown in , the calculation principle of the formula for human height h is as follows, where Figure 3 The human body at position 1 on the left is on the left side of the central axis of the binocular camera (dashed arrow), and the height coordinate value Y1 is positive. The human body at position 2 on the right is on the right side of the central axis of the binocular camera, and the height coordinate value Y2 is negative.

[0135] After obtaining the depth coordinate value and height coordinate value of the target point of the face portrait, for the human body at position 1 (face portrait No. 1) in the figure, auxiliary lines OC1, A1C1, A1B1, B1C1, etc. are drawn in the figure, where point B1 is the target point of face portrait No. 1, point O is located in the plane where the binocular camera is located (i.e., the plane formed by the X and Y coordinate axes) and is located on the central axis of the binocular camera or on a line parallel to the central axis. In this embodiment, O is the coordinate origin of the plane where the Y and Z axes are located, and C1 is the coordinate origin of B1 in the plane formed by the X and Y coordinate axes. The projection points of the plane; B1C1=Z1; Point A1 is the intersection of the first straight line passing through point C1 and perpendicular to the horizontal plane and the second straight line passing through point B1 and parallel to the horizontal plane, wherein the first straight line and the second straight line are perpendicular; Point D1 is the intersection of the third straight line passing through point O and perpendicular to the horizontal plane and the fourth straight line passing through point C1 and parallel to the horizontal plane, and the third straight line and the fourth straight line are perpendicular; Point O1 is the intersection of the fifth straight line passing through point O and parallel to the horizontal plane and the sixth straight line passing through B1 and perpendicular to the horizontal plane, and the fifth straight line and the sixth straight line are perpendicular.

[0136] Based on the above auxiliary line calculation Figure 1 The vector length of the midline segment A1C1 is:

[0137] A1C1=Z1cosα;

[0138] calculate Figure 1 The length of the midline segment OD1 vector is:

[0139] OD1 = Y1sinα;

[0140] Based on this, the distance |O1B1| from the face avatar target point i to the plane where the binocular camera is installed can be calculated as:

[0141] |O1B1|=Z1cosα+Y1sinα;

[0142] Since the binocular camera is installed at a height of M, the human height h can be obtained as:

[0143] h=M-(Z1cosα+Y1sinα);

[0144] Based on the above principle, this embodiment can realize the detection of human height.

[0145] Similarly, for Figure 3 For the human body at position No. 2 (face avatar No. 2), auxiliary lines B2C2, B2A2, B2O2, A2O2, etc. are drawn in the figure, among which point B2 is the target point of face avatar No. 2, C2 is the projection point of B2 on the plane formed by the X and Y coordinate axes; point A2 is the intersection of a straight line passing through point C2 and parallel to the horizontal plane and a straight line passing through point B2 and perpendicular to the horizontal plane, and the above two straight lines are perpendicular; point O2 is the intersection of a straight line passing through point O and parallel to the horizontal plane and a straight line passing through point B2 and perpendicular to the horizontal plane, and the above two straight lines are also perpendicular.

[0146] Based on the above auxiliary line calculation Figure 1 The vector length of the midline segment B2A2 is:

[0147] B2A2=Z2cosα;

[0148] calculate Figure 1 The length of the midline segment A2O2 vector is:

[0149] A2O2=Y2sinα;

[0150] Based on this, the distance |O2B2| from the face avatar target point i to the plane where the binocular camera is installed can be calculated as:

[0151] |O2B2|=Z2cosα+Y2sinα; where Y2 is a negative value.

[0152] Since the binocular camera is installed at a height of M, the human height h can be obtained as:

[0153] h=M-(Z2cosα+Y2sinα).

[0154] The principle of the above-mentioned human height calculation formula in this embodiment can be used to deduce that the height detection in this embodiment is achievable and accurate.

[0155] After step S203 of the present embodiment, considering that the height of the target point i of the face avatar is calculated in step S203, and the point may be the grayscale center of gravity or the center point of the face avatar, in order to make the height of the human body detected by the method of the present embodiment more accurate, a compensation process of the height of the human body is performed in step S204, which is as follows:

[0156] S204, for the face avatar area extracted from the image information, determine the pixel point j at the highest position of the face avatar in the image, then calculate the distance h" between the face avatar target point i and the pixel point j in the image, and add the h value to h" as the final human height detection result H; wherein:

[0157]

[0158] H=h+h".

[0159] (x i ,y i ) is the coordinate of the target point i in the face portrait in the image (left image or right image), (x j ,y j ) is the coordinate of the pixel point j at the highest position of the face portrait in the image (left image or right image). The above coordinates are all the coordinates of the pixel points on the left image or the right image, and do not need to be obtained through depth calculation. In this embodiment, the pixel point j at the highest position of the face portrait can be a point corresponding to the top of the head of the face portrait, or a point on the upper border of the face portrait box. If the pixel point j is a point on the upper border of the face portrait box, then the pixel point j is the projection of the target point i on the upper border of the face portrait box, that is, the line connecting the pixel points i and j is perpendicular to the upper border of the face portrait box.

[0160] In this embodiment, h" may also be set as a fixed compensation value, and the human height h value calculated in step S203 is added to the fixed compensation value h" as the final human height detection result. In this embodiment, if the target point i of the face avatar is the center of gravity or center of the face avatar, the distance from the target point of the face avatar of multiple people to the highest position of the face avatar can be obtained, and the average is taken as the fixed compensation value.

[0161] In this embodiment, when the gesture action is to control the clothes drying machine to perform the function of lowering the clothes drying rod, the clothes drying machine clothes drying rod lowering stroke can be controlled according to the human height detected in the above human height detection step. In this embodiment, corresponding clothes drying rod lowering strokes can be configured for different human heights, for example, as follows:

[0162] If the height of the person is less than 165cm, the lowering stroke of the clothes drying rod is 100%, for example, the corresponding lowering height can be 120cm;

[0163] The height of a person is 165cm to 170cm, and the descending stroke of the drying pole is 96%, for example, the corresponding descending height can be 115cm;

[0164] The height of a person is 170cm to 175cm, and the descending stroke of the clothes-drying rod is 92%, for example, the corresponding descending height can be 110cm;

[0165] The height of a person is 175cm to 180cm, and the descending stroke of the clothes-drying rod is 88%, for example, the corresponding descending height can be 105cm;

[0166] The height of a person is 180cm to 185cm, and the descending stroke of the drying pole is 83%, for example, the corresponding descending height can be 100cm;

[0167] The height of a person is 185cm to 190cm, and the descending stroke of the drying pole is 79%, for example, the corresponding descending height can be 95cm;

[0168] The height of a person is 190cm to 195cm, and the descending stroke of the clothes-drying rod is 75%, for example, the corresponding descending height can be 90cm;

[0169] The height of a person is 195cm to 200cm, and the descending stroke of the drying pole is 71%, for example, the corresponding descending height can be 85cm;

[0170] When the height of a person is higher than 200 cm, the descending stroke of the clothes drying pole is 66%, for example, the corresponding descending height may be 80 cm.

[0171] The above is only one of the data configuration methods. Specifically, the corresponding pole lowering stroke, i.e., the corresponding lowering height, can be configured for different human heights according to the actual installation environment of the clothes drying machine (e.g., the height of the balcony) or user needs. Based on the above steps, this embodiment can adaptively adjust the lowering height of the pole according to the height of the human body, making it convenient for different users to dry and collect clothes.

[0172] like Figure 4 As shown in , the following specifically illustrates one complete implementation process of the method for controlling a clothes drying machine based on gestures in this embodiment:

[0173] S1. The human body sensing unit installed on the clothes drying machine monitors whether there is anyone entering the drying sensing area. If so, the binocular camera is started to collect images, and then the process proceeds to step S2.

[0174] S2. Obtain a left image and a right image captured by a binocular camera, crop the left image and the right image respectively, and normalize the cropped left image and the right image.

[0175] S3. Detect whether there is a face in the normalized left image. If face information is detected, go to step S4; otherwise, return to step S1.

[0176] S4. For the left image normalized in step S3, extract the target gesture from it, classify it, and determine the gesture action. The detailed processing process is shown in the above steps S102 and S103.

[0177] S5, according to the determined gesture action, control the clothes drying machine to perform a corresponding function. If the gesture action matches the control of the clothes drying machine's drying rod to descend, then execute step S6.

[0178] S6. Detect whether there is a face in the normalized right image. The detailed process is shown in the above step S201. If there is a face, execute step S7, otherwise return to step S2.

[0179] S7. Obtain a left face portrait and a right face image for the normalized left and right images, calculate depth maps for the left face portrait and the right face image, and calculate the height of the human body according to the depth map calculation results; see the above steps S202 and S203 for the detailed process.

[0180] S8. The height of the human body calculated in step S7 controls the corresponding descending stroke of the clothes drying rod of the clothes drying machine.

[0181] In this embodiment, the above steps S1 to S8 are only one of the implementation processes. In the actual control process of the clothes drying machine, when the gesture action matches other functions, steps S6 to S8 do not need to be executed. Or when the gesture action matches the lowering of the clothes drying rod, the height information of the human body can also be obtained based on the method of steps Sa to Sc, and the clothes drying rod can be controlled to lower to the corresponding height based on this.

[0182] In summary, the present embodiment controls the clothes drying machine through gestures, specifically, gesture recognition is initiated when facial information is detected. Since the image acquisition device installed on the clothes drying machine is located above the user's head, the user needs to consciously look up to collect facial information. Therefore, it can avoid the situation where users who enter the drying sensing area without control awareness accidentally trigger the clothes drying machine due to coincidental gestures, which can effectively improve the accuracy of clothes drying machine recognition and avoid false triggering. Furthermore, the method of the present embodiment adds human height detection. When the gesture action matches the function of lowering the clothes drying rod, the corresponding stroke of the clothes drying rod lowering can be controlled based on the human height detection result; based on this, the method of the present embodiment can control the clothes drying rod to drop to a suitable height for each user according to the user's human height, and realize adaptive adjustment of the height of the clothes drying rod lowering, which is convenient for different users to dry and collect clothes.

[0183] Those skilled in the art will appreciate that all or part of the steps in the method of the present embodiment can be completed by instructing the relevant hardware through a program, and the corresponding program can be stored in a computer-readable storage medium. It should be noted that although the method operation of the present embodiment is described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in this specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the steps depicted can change the order of execution, some steps can also be performed simultaneously, additionally or alternatively, some steps can be omitted, multiple steps can be combined into one step for execution, and / or one step can be decomposed into multiple steps for execution.

[0184] This embodiment 2

[0185] This embodiment discloses a clothes drying machine control system, which is applied to the clothes drying machine. Figure 5As shown in , it includes a controller, an image processor, an image acquisition unit and a human body sensing unit. The controller can be a controller in the clothes drying machine host, or another controller set separately. If it is the latter, the controller is connected to the clothes drying machine main control board and exchanges information with the clothes drying machine main control board. Among them:

[0186] In this embodiment, the human body sensing unit is connected to the controller to detect whether a person enters the drying sensing area. When a person enters, the image acquisition unit can be activated to start image acquisition. The human body sensing unit can be a sensor that can sense the human body, such as an infrared sensor.

[0187] In this embodiment, the image acquisition unit is connected to the image processor to acquire image information of the drying sensing area, and the acquired image information is sent to the image processor for image processing. In this embodiment, the image acquisition unit uses a binocular camera, which can be installed on the main body shell of the clothes drying machine or embedded in the main body shell. In this embodiment, the binocular camera is installed at an angle, and the inclination angle can be 10° to 85°. The inclination angle refers to the angle between the central axis of the binocular camera and the plumb line, which can be specifically as follows Figure 3 In this embodiment, the preferred tilt angle is 30° or about 30°. When the tilt angle is 30°, the drying sensing area of ​​the binocular camera can be a rectangular area of ​​2m*1.2m. The binocular camera installed at a certain tilt angle can make it possible for a person who enters the drying sensing area to have his / her face image captured only when he / she looks up at the image acquisition unit. This can effectively avoid false triggering when there is no intention to dry the clothes.

[0188] In this embodiment, the binocular camera is installed on the side of the clothes drying machine main body shell. For the clothes drying machine installed on the balcony, the shooting direction of the binocular camera is toward the side entering the balcony, which is usually located on the side of the long side of the clothes drying machine main body shell. The tilted binocular camera ensures that an area on the side of the clothes drying machine is a drying sensing area. Objects entering this area will be photographed by the binocular camera, including faces and gestures. In addition, the installation position of the above-mentioned human body sensing unit can also be consistent with the image acquisition unit. The sensing range of the human body sensing unit is greater than or equal to the drying sensing area of ​​the image acquisition unit. It is best that the two can be as consistent as possible. The first camera and the second camera in the binocular camera respectively capture the left image and the right image as the collected image information of the drying sensing area.

[0189] In this embodiment, the image processor is connected to the image acquisition unit, and is used to receive the image information collected by the image acquisition unit and perform image processing. In this embodiment, the image processor receives the image information collected by the image acquisition unit, that is, receives the left image and the right image captured by the first camera and the second camera of the binocular camera, and performs image processing. The image processing process includes the face avatar extraction detection, target gesture extraction and gesture action recognition of steps S101 to S103 in embodiment 1, and also includes the human body height detection step in embodiment 1. After the image processor recognizes the gesture action from the image information, the recognized gesture action information is sent to the controller; at the same time, if the human body height detection step is performed, the detected human body height information can be sent to the controller, or the detected human body height information can be sent to the controller only when the recognized gesture action matches the clothes drying machine descending function.

[0190] In this embodiment, the controller is connected to the image processor, receives the gesture action information sent by the image processor, and controls the clothes drying machine to perform the corresponding function based on the clothes drying machine function matched by the gesture action information. For example, if the first gesture action corresponds to turning on the lighting, the controller controls the clothes drying machine lighting to turn on after receiving the first gesture action. In addition, after the controller receives the human height information sent by the image processor, if the currently received gesture action matches the clothes drying machine descending function, the stroke of the clothes drying rod descending is controlled according to the received human height information, wherein the purpose of controlling the stroke of the clothes drying rod descending can be achieved by controlling the number of revolutions of the motor in the clothes drying machine, or by other methods such as sensors to control the stroke of the clothes drying rod descending, which is not limited here. In addition, if a special gesture action is configured for the user as shown in Example 1, the controller obtains the human height information configured by the special gesture action after receiving the special gesture action, and controls the corresponding stroke of the clothes drying rod descending according to the human height information; wherein the human height information configured by the special gesture action can be stored in the controller in advance.

[0191] In this embodiment, the image acquisition unit (such as a binocular camera), the image processor and the human body sensing unit can be arranged in a structural member, for example, all arranged in a shell, and the shell is installed on one side of the long side of the clothes drying machine main body shell. After the shell is installed, the angle between the central axis of the binocular camera image and the plumb line, that is, the tilt angle, reaches 10° to 85°; the corresponding connecting wires of the image processor, including the power connecting wire and the signal connecting wire, pass through the shell and the clothes drying machine main body shell, and are connected to the controller in the clothes drying machine main body. In addition, a controller can also be arranged in the shell, and the corresponding connecting wires of the controller pass through the shell and the clothes drying machine main body shell, and are connected to the main control board in the clothes drying machine main body.

[0192] Example 3

[0193] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the gesture-based clothes drying machine control method described in Embodiment 1, including:

[0194] Acquiring image information acquired by an image acquisition unit;

[0195] For image information, detect whether there is face information;

[0196] If there is face information in the image information, extract the target gesture from the image information;

[0197] Identify the extracted target gesture and determine the gesture action;

[0198] The clothes drying machine is controlled to execute corresponding functions according to the determined gesture actions.

[0199] The specific implementation process of the above steps is shown in Example 1 and will not be repeated here.

[0200] In this embodiment, the electronic device can be a device with the above-mentioned functions set on a clothes drying machine, including: a processor, a memory, a bus and a communication interface, the processor, the communication interface and the memory are connected through the bus; the processor is configured to execute an executable module stored in the memory, such as a computer program.

[0201] Example 4

[0202] This embodiment discloses a storage medium storing a program. When the program is executed by a processor, the method for controlling a clothes drying machine based on gestures described in Embodiment 1 is implemented, including:

[0203] Acquiring image information acquired by an image acquisition unit;

[0204] For image information, detect whether there is face information;

[0205] If there is face information in the image information, extract the target gesture from the image information;

[0206] Identify the extracted target gesture and determine the gesture action;

[0207] The clothes drying machine is controlled to execute corresponding functions according to the determined gesture actions.

[0208] The specific implementation process of the above steps is shown in Example 1, which will not be repeated here. In this embodiment, the storage medium can be a disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), a USB flash drive, a mobile hard disk, and other media.

[0209] The above is only an optional implementation method for some implementation scenarios of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the solution of the present application, other similar implementation methods based on the technical ideas of the present application are also within the protection scope of the embodiments of the present application.

Claims

1. A method for controlling a clothes drying machine based on gestures, characterized in that: The method is applied to a clothes drying machine equipped with an image acquisition unit, and comprises the following steps: Acquiring image information acquired by an image acquisition unit; For image information, detect whether there is face information; If there is face information in the image information, extract the target gesture from the image information; Identify the extracted target gesture and determine the gesture action; The clothes drying machine is controlled to execute corresponding functions according to the determined gesture actions.

2. The method for controlling a clothes drying machine based on gestures according to claim 1, characterized in that: For the image information collected by the image acquisition unit, when multiple gestures are extracted, the distance between each gesture and the face is calculated based on each gesture area and the face avatar area in the image information, and the gesture closest to the face is selected as the target gesture.

3. The method for controlling a clothes drying machine based on gestures according to claim 2, characterized in that: The distance between each gesture and the face is calculated in the following way: Among them, (x i ,y i ) is the coordinate of the target point i in the face image in the image, (x s ,y s ) is the coordinate of the gesture target point s in the image; D is is the distance between the gesture target point s and the face avatar target point i in the image.

4. The method for controlling a clothes drying machine based on gestures according to claim 1, characterized in that: The following steps are also included: Setting a dedicated gesture action; wherein the dedicated gesture action is used to control the clothes drying machine to execute the function of lowering the clothes drying rod; For special gesture actions, configure human height information; When the gesture action in the image information is recognized as a special gesture action, the configured human body height information is obtained, and the clothes drying rod of the clothes drying machine is controlled to descend to a corresponding stroke according to the human body height information.

5. The method for controlling a clothes drying machine based on gestures according to claim 1, characterized in that: The following steps of human height detection are also included: For the image information collected by the image acquisition unit, extract the face image from the image information, and perform depth map calculation to obtain a depth map calculation result; Calculate the height h of the human body based on the depth map calculation result: h=M-(Z i cosα+Y i sinα); in: M is the installation height of the image acquisition unit on the clothes drying machine, α is the inclination angle of the image acquisition unit installation, corresponding to the angle between the central axis of the image acquisition unit and the plumb line; Z i and Y i is the depth map calculation result, Z i is the coordinate value of the target point i of the face portrait at depth, Y i is the height coordinate value of target point i of the face portrait; When the gesture action is to control the clothes drying machine to execute the function of lowering the clothes drying rod, the clothes drying machine clothes drying rod is controlled to lower the corresponding stroke according to the human body height detected in the above-mentioned human body height detection step.

6. The method for controlling a clothes drying machine based on gestures according to claim 5, characterized in that: The image acquisition unit is a binocular camera installed on the clothes drying machine at a set angle, including a first camera and a second camera; a left image and a right image acquired by the first camera and the second camera are acquired respectively; Detecting facial information, determining gestures, and obtaining depth map calculation results by any of the following methods: The first one: The left image and the right image are cropped respectively; For the cropped left image, normalize it and detect the face information; When face information is detected in the left image, the target gesture is extracted from the normalized left image; Classify and identify the extracted target gestures to determine the gesture actions; For the cropped right image, normalize it and detect the face information; When face information is detected in the right image, extract the left face image and the right face image respectively for the left image and the right image; For the face portrait in the left image and the face portrait in the right image, depth map calculation is performed to obtain depth map calculation results; Second type: The left image and the right image are cropped respectively; For the cropped right image, normalize it and detect the face information; When face information is detected in the right image, target gesture features are extracted from the normalized right image; Classify and identify the extracted target gesture features to determine the gesture action; For the cropped left image, normalize it and detect the face information; When face information is detected in the left image, extracting a left face avatar and a right face avatar from the left image and the right image respectively; A depth map calculation is performed for the left-hand face portrait and the right-hand face portrait to obtain a depth map calculation result.

7. The method for controlling a clothes drying machine based on gestures according to claim 5, characterized in that: In the human height detection step: When a human face image is detected in the image information, the human face image is used as the target human face image, and a depth map calculation is performed to obtain a depth map calculation result; When multiple face images are extracted from the image information, the target face image is determined by any of the following methods, and a depth map is calculated for the target face image to obtain a depth map calculation result: The first one: Obtaining the coordinate value of each face avatar target point in the image information at depth, and determining the depth information of each face avatar through the coordinate value of each face avatar target point at depth; then comparing the depth information of each face avatar, and determining the target face avatar based on the face avatar with the smallest depth distance; Second type: Obtain the coordinate value of the height of each face avatar target point in the image information, and determine the height information of each face avatar through the coordinate value of the height of each face avatar target point; then compare the height information of each face avatar, and determine the target face avatar based on the face avatar with the smallest height distance; The third type: For each face avatar extracted from the image information, a depth map is calculated respectively, and then the corresponding human height is detected through the human height detection step; the human heights corresponding to each face avatar are compared, and the target face avatar is determined based on the face avatar with the shortest human height.

8. The method for controlling a clothes drying machine based on gestures according to claim 5, characterized in that: When the image acquisition unit is a binocular camera, in the human height detection step: obtain the left image and the right image acquired by the binocular camera; for the left image and the right image, determine whether the face portraits extracted therein meet the following conditions: in the pixel coordinates (x, y) of the top corner of the face frame, |y| is less than a set value; if so, determine the extracted face portrait as the target face image; otherwise, determine that there is no target face image in the image; In the human body height detection step, a depth map is calculated for the target face portrait to obtain a depth map calculation result.

9. The method for controlling a clothes drying machine based on gestures according to claim 5, characterized in that: The human body height detection step also includes a human body height compensation step, which is achieved by any of the following: For the face image area extracted from the image information, determine the pixel point j at the highest position of the face image in the image, then calculate the distance h" between the face image target point i and the pixel point j in the image, and add the h value to h" as the final human height detection result; Set h" as a fixed compensation value, and add the h value to h" as the final human height detection result.

10. A clothes drying machine control system, characterized in that: It includes a controller, an image processor and an image acquisition unit; The image acquisition unit is used to acquire image information; The image processor is connected to the image acquisition unit and is used to detect facial information from the image information acquired by the image acquisition unit, extract target gesture features from the image information when it is determined that facial information is detected in the image information, and identify the extracted target gesture features to determine the gesture action; The controller is connected to the image processor and is used to obtain the gesture action determined by the image processor, and control the clothes drying machine to perform corresponding functions according to the determined gesture action.

11. The clothes drying machine control system according to claim 10, characterized in that: It also includes a human body sensing unit; The human body sensing unit is connected to the controller and is used to detect whether a person enters the drying sensing area. When a person enters, the controller starts the image acquisition unit to collect image information.

12. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the gesture-based clothes drying machine control method according to any one of claims 1 to 9 is implemented.

13. An electronic device comprising a processor and a memory for storing a program executable by the processor, characterized in that: When the processor executes the program stored in the memory, the gesture-based clothes drying machine control method according to any one of claims 1 to 9 is implemented.

Images