Touch control for a range hood

Abstract

The application discloses a touch type range hood controller and relates to the technical field of human-computer interaction, which comprises an integrated operation panel, a control module and a data acquisition module, wherein the data acquisition module comprises a position acquisition module and a touch screen acquisition module; the position acquisition module acquires a position range where a user is located; when it is detected that the user reaches a preset position range, the touch screen acquisition module is started; the touch screen acquisition module determines a space vector according to a touch screen area of the integrated operation panel contacted by the user; the space vector is used to determine icons of corresponding keys, and each icon corresponds to a touch screen area; the control module prestores key information corresponding to the touch screen area, and starts corresponding range hood keys according to the touch screen area. The touch type range hood controller disclosed by the application solves the problem that the operation panel of the range hood fails due to fingers being stained with water or stains when cooking in the kitchen.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction technology, and more specifically to a touch-screen range hood controller. Background Technology

[0002] Traditional range hoods use mechanical switches, which often become clogged with grease after a period of use, affecting operation and making cleaning inconvenient. In recent years, most range hoods on the market have used touch switches. Touch switches have no mechanical action and are insulated, dustproof, waterproof, and oilproof, preventing grease buildup. However, when cooking in the kitchen, people often have water, oil, or flour on their hands. If the touch switch is used to turn on the range hood under these conditions, it frequently malfunctions or misoperates. Furthermore, the grease on the switch surface will soil hands, affecting continued cooking.

[0003] Therefore, how to solve the problem of touch switches malfunctioning due to deposits is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] In view of this, the present invention provides a touch-sensitive range hood controller. By integrating the control panel and establishing a three-dimensional coordinate system to identify the position of the control panel touched by the operator's finger, the corresponding range hood switch is turned on. This solves the technical problem in the prior art where the touch switch malfunctions or misoperates when the operator's hands are covered with water, oil or flour while cooking in the kitchen.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A touch-sensitive range hood controller includes: an integrated operation panel, a control module, and a data acquisition module. The data acquisition module includes a position acquisition module and a touch screen acquisition module. The position acquisition module acquires the user's location range, and when the user is detected to have reached a preset position range, the touch screen acquisition module is activated. The touch screen acquisition module determines a spatial vector based on the touch screen area of ​​the integrated operation panel touched by the user. The spatial vector is used to determine the icons of the corresponding buttons, with each icon corresponding to a touch screen area. The control module pre-stores the button information corresponding to the touch screen areas and activates the corresponding range hood buttons based on the touch screen areas.

[0007] Preferably, it also includes a pollution detection module, which detects pollution in the data acquisition module, sets a pollution threshold, and issues an alarm when the pollution in the data acquisition module reaches the pollution threshold.

[0008] Preferably, the integrated operation panel is provided with a contact area and a non-contact area, the contact area is provided with icons for corresponding buttons, and the non-contact area is provided with a data acquisition module.

[0009] Preferably, the location acquisition module specifically includes:

[0010] The image acquisition module acquires a target ranging image;

[0011] The target bounding box acquisition module inputs the target ranging image into a pre-trained deep learning convolutional network model for target detection, identifies the user in the image, and acquires the target bounding box.

[0012] The knowledge acquisition module collects and calibrates the user's RGB image to obtain prior knowledge of the relationship between distance and pixel height difference;

[0013] The range detection module has a preset location range. It uses prior knowledge from the knowledge acquisition module to estimate the distance to the user in the target box, obtains the estimated distance, and compares the preset location range with the estimated distance. If the estimated distance is less than or equal to the preset location range, the touch screen acquisition module is activated.

[0014] Preferably, the touch screen acquisition module specifically includes: a finger calibration module, an image acquisition module, and a processing module. The finger calibration module calibrates the user's fingertip as a marker point, the image acquisition module acquires the coordinates of the marker point and sends them to the processing module, and the processing module determines the touch screen area contacted by the marker point.

[0015] Preferably, the hand calibration module specifically includes:

[0016] The teacher feature acquisition module acquires the user's hand image, and after data augmentation processing, inputs the user's hand image into the teacher network for forward computation to obtain convolutional feature maps with different channel feature information.

[0017] The refocusing matrix module aggregates the convolutional feature map to obtain a channel information aggregation map, and generates a region refocusing matrix based on the position coordinates of the fingertip region and the channel information aggregation map.

[0018] The student feature acquisition module performs the same data augmentation process on the user's hand image and then simultaneously inputs it into the student network for forward computation to obtain a convolutional feature map.

[0019] The convergence module uses the region refocusing matrix combined with the convolutional feature maps of the teacher network and the student network to perform scaling penalty training on the student network until the object detection accuracy of the student network converges.

[0020] The marking module acquires an image of the user's hand within a fixed range near the integrated operation panel, inputs it into the student network, identifies the fingertips, and marks them to obtain marker points.

[0021] Preferably, the image acquisition module specifically includes: a first image acquisition module and a second image acquisition module, wherein the first image acquisition module and the second image acquisition module respectively acquire images of the marker points at a preset step size and transmit the acquired images to the processing module.

[0022] Preferably, the processing module specifically includes: a coordinate acquisition module, a marker point coordinate system establishment module, and a region determination module;

[0023] The coordinate acquisition module performs feature point matching on the first and second marked images acquired during the initial image acquisition by the first and second image acquisition modules to obtain the feature point coordinates 'a' in the coordinate system of the first image acquisition module. R The coordinates of the corresponding feature point a in the coordinate system of the second image acquisition module L ;

[0024] In world coordinates, calculate the unit direction vectors N1 and N2 from the first and second image acquisition modules to the matching point, respectively:

[0025]

[0026]

[0027] Among them, Z R Z L R represents the coefficients that normalize the direction vectors of the first image acquisition module and the second image acquisition module, respectively. RW R LW K1 and K2 represent the rotation matrices of the coordinate systems of the first and second image acquisition modules relative to the world coordinate system, respectively. K1 and K2 represent the intrinsic parameters of the first and second image acquisition modules, respectively. The superscript T indicates transpose.

[0028] Calculate the spatial coordinates A of the first marked point:

[0029]

[0030]

[0031]

[0032] Where b = N1 T N2, A 10 A 20 These represent the first image acquisition module O. RSecond image acquisition module O L In the world coordinate system, A1 and A2 are the spatial coordinates of ray O. RaR and O LaL The foot of the perpendicular to the common perpendicular;

[0033] The marker point coordinate system establishment module establishes the marker point coordinate system with the spatial coordinates A of the first marker point as the origin of the marker point coordinate system, and performs marker point image acquisition with the preset step size to obtain the spatial coordinates B of the last acquired marker point and obtain the spatial vector AB.

[0034] The area determination module divides the integrated operation panel into blocks according to the touch screen area corresponding to each icon, and inputs the coordinate system of the marked point. Based on the touch screen area of ​​the icon through which the spatial vector AB passes, it sends the corresponding range hood button to the control module.

[0035] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a touch-sensitive range hood controller. Compared with the prior art, by integrating the control panel and establishing a three-dimensional coordinate system to identify the position of the control panel touched by the operator's finger, the corresponding range hood switch is turned on. This solves the problem of touch switch malfunction or erroneous operation in the prior art when the operator's hands are covered with water, oil or flour while cooking in the kitchen. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0037] Figure 1 The attached figure is a schematic diagram of the structure provided by the present invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] This invention discloses a touch-sensitive range hood controller, such as... Figure 1As shown, it includes: an integrated operation panel, a control module, and a data acquisition module. The data acquisition module includes a location acquisition module and a touch screen acquisition module. The location acquisition module acquires the user's location range. When the user is detected to have reached the preset location range, the touch screen acquisition module is activated. The touch screen acquisition module determines the spatial vector based on the touch screen area of ​​the integrated operation panel that the user touches. The spatial vector is used to determine the icon of the corresponding button, and each icon corresponds to a touch screen area. The control module pre-stores the button information corresponding to the touch screen area and activates the corresponding range hood button based on the touch screen area.

[0040] In one specific embodiment, the integrated control panel can be made of glass.

[0041] In one specific embodiment, a pollution detection module is also included. The pollution detection module detects pollution in the data acquisition module, sets a pollution threshold, and issues an alarm when the pollution in the data acquisition module reaches the pollution threshold.

[0042] Preferably, the integrated operation panel is provided with a contact area and a non-contact area. The contact area is provided with icons for the corresponding buttons, and the non-contact area is provided with a data acquisition module.

[0043] Preferably, the location acquisition module specifically includes:

[0044] The image acquisition module acquires a target ranging image;

[0045] The target bounding box acquisition module inputs the target ranging image into a pre-trained deep learning convolutional network model for target detection, identifies the user in the image, and acquires the target bounding box.

[0046] The knowledge acquisition module collects and calibrates the user's RGB image to obtain prior knowledge of the relationship between distance and pixel height difference;

[0047] The range detection module has a preset location range. It uses prior knowledge from the knowledge acquisition module to estimate the distance to the user in the target box and obtains the estimated distance. The preset location range is compared with the estimated distance. If the estimated distance is less than or equal to the preset location range, the touch screen acquisition module is activated.

[0048] The specific method for estimating the distance between the user and the RGB camera using a projection inversion method based on prior knowledge to estimate the distance to the target in the candidate box is as follows:

[0049] When the user's height is fixed, we can deduce from similar triangles that:

[0050]

[0051]

[0052]

[0053] D1 = D2 × probability × a

[0054] In the formula, a is a parameter obtained through prior knowledge, D2 is the camera focal length, h1 is the user's height in the scene, H1 is the maximum height of the scene captured by the camera, D1 is the distance between the user and the lens, h2 is the height of the user projected onto the photo, and H2 is the maximum height of the scene projected onto the photo.

[0055] Let D2×a=γ, then as long as γ and propotion are known, the distance between the user and the lens can be calculated.

[0056] In one specific embodiment, the touch screen acquisition module specifically includes: a finger calibration module, an image acquisition module, and a processing module. The finger calibration module calibrates the user's fingertip as a marker point, the image acquisition module acquires the coordinates of the marker point and sends them to the processing module, and the processing module determines the touch screen area touched by the marker point.

[0057] In one specific embodiment, the hand index calibration module specifically includes:

[0058] The teacher feature acquisition module acquires user hand images, and after data augmentation, inputs the user hand images into the teacher network for forward computation to obtain convolutional feature maps with different channel feature information; the teacher network is a convolutional neural network capable of independently performing hand image detection;

[0059] The refocusing matrix module aggregates the convolutional feature maps to obtain a channel information aggregation map. Based on the location coordinates of the fingertip region and the channel information aggregation map, a region refocusing matrix is ​​generated.

[0060] The student feature acquisition module performs the same data augmentation process on the user's hand image and then inputs it into the student network for forward computation to obtain a convolutional feature map; the student network is a convolutional neural network with a smaller scale than the teacher network.

[0061] The convergence module uses a region refocusing matrix combined with the convolutional feature maps of the teacher and student networks to perform scaling penalty training on the student network until the object detection accuracy of the student network converges.

[0062] The marking module acquires an image of the user's hand within a fixed range near the integrated control panel, inputs it into the student network, identifies the fingertips, and marks them to obtain marker points.

[0063] In one specific embodiment, the image acquisition module specifically includes: a first image acquisition module and a second image acquisition module. The first image acquisition module and the second image acquisition module respectively perform image acquisition of the marker points at a preset step size and transmit the acquired images to the processing module.

[0064] In one specific embodiment, the processing module specifically includes: a coordinate acquisition module, a marker point coordinate system establishment module, and a region determination module;

[0065] The coordinate acquisition module performs feature point matching on the first and second marked images acquired during the initial image acquisition by the first and second image acquisition modules to obtain the feature point coordinates 'a' in the coordinate system of the first image acquisition module. R The coordinates of the corresponding feature point a in the coordinate system of the second image acquisition module L ;

[0066] In world coordinates, calculate the unit direction vectors N1 and N2 from the first and second image acquisition modules to the matching point, respectively:

[0067]

[0068]

[0069] Among them, Z R Z L R represents the coefficients that normalize the direction vectors of the first image acquisition module and the second image acquisition module, respectively. RW R LW K1 and K2 represent the rotation matrices of the coordinate systems of the first and second image acquisition modules relative to the world coordinate system, respectively. K1 and K2 represent the intrinsic parameters of the first and second image acquisition modules, respectively. The superscript T indicates transpose.

[0070] Calculate the spatial coordinates A of the first marked point:

[0071]

[0072]

[0073]

[0074] Where b = N1 T N2, A 10 A 20 These represent the first image acquisition module O. R Second image acquisition module O L In the world coordinate system, A1 and A2 are the spatial coordinates of ray O. RaR and O LaL The foot of the perpendicular to the common perpendicular;

[0075] The marker point coordinate system establishment module establishes the marker point coordinate system with the spatial coordinates A of the first marker point as the origin of the marker point coordinate system, and performs marker point image acquisition with a preset step size to obtain the spatial coordinates B of the last acquired marker point and obtain the spatial vector AB.

[0076] In one specific embodiment, since both the first image acquisition module and the second image acquisition module are located in fixed positions within the non-contact area of ​​the integrated operation panel, according to the first image acquisition module O R Second image acquisition module O L In the world coordinate system, the integrated operation panel is transformed into a plane corresponding to spatial coordinate A, and the corresponding contact area in the plane is divided into blocks.

[0077] The area determination module divides the integrated operation panel into blocks according to the touch area corresponding to each icon, and inputs the coordinate system of the marked point. Based on the touch area of ​​the icon that the spatial vector AB passes through, it sends the signal to the control module to turn on the corresponding range hood button.

[0078] This invention uses a location acquisition module to collect the user's location. When the user approaches a designated area, the touchscreen acquisition module is activated. The finger calibration module in the touchscreen acquisition module marks the user's fingertip as a marker point. The first and second image acquisition modules in the image acquisition module collect images of the marker point gradually approaching the integrated operation panel according to a preset step size, and send them to the processing module. The processing module obtains the position of the first occurrence of the marker point based on the positions of the first and second image acquisition modules, uses it as the origin of the spatial coordinate system, and obtains the position of the last occurrence of the marker point. The first and last occurrence positions of the marker point are connected in the spatial coordinate system to obtain the corresponding vector. At the same time, it is input into the plane of the integrated operation panel area determined by the first and second image acquisition modules to divide the area corresponding to the button icons of the integrated operation panel. When the corresponding vector passes through the area of ​​a certain button icon, the control module activates the corresponding range hood button.

[0079] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0080] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A touch control for a range hood, comprising: include: The system includes an integrated operation panel, control module, and data acquisition module. The data acquisition module includes a location acquisition module and a touch screen acquisition module. The location acquisition module acquires the user's location range, and the touch screen acquisition module is activated when the user is detected to have reached a preset location range. The touchscreen acquisition module determines a spatial vector based on the touchscreen area where the user touches the integrated operation panel; it uses the spatial vector to determine the icon of the corresponding button, with each icon corresponding to a touchscreen area; the control module pre-stores the button information corresponding to the touchscreen area, and activates the corresponding range hood button based on the touchscreen area. The location acquisition module specifically includes: The image acquisition module acquires a target ranging image; The target bounding box acquisition module inputs the target ranging image into a pre-trained deep learning convolutional network model for target detection, identifies the user in the image, and acquires the target bounding box. The knowledge acquisition module collects and calibrates the user's RGB image to obtain prior knowledge of the relationship between distance and pixel height difference; The range detection module has a preset position range. It uses prior knowledge from the knowledge acquisition module to estimate the distance to the user in the target box, obtains the estimated distance, and compares the preset position range with the estimated distance. If the estimated distance is less than or equal to the preset position range, the touch screen acquisition module is activated. The touchscreen acquisition module specifically includes: a finger calibration module, an image acquisition module, and a processing module. The finger calibration module calibrates the user's fingertip as a marker point. The image acquisition module acquires the coordinates of the marker point and sends them to the processing module. The processing module determines the touchscreen area touched by the marker point. The manual index calibration module specifically includes: The teacher feature acquisition module acquires the user's hand image, and after data augmentation processing, inputs the user's hand image into the teacher network for forward computation to obtain convolutional feature maps with different channel feature information. The refocusing matrix module aggregates the convolutional feature map to obtain a channel information aggregation map, and generates a region refocusing matrix based on the position coordinates of the fingertip region and the channel information aggregation map. The student feature acquisition module performs the same data augmentation process on the user's hand image and then simultaneously inputs it into the student network for forward computation to obtain a convolutional feature map. The convergence module uses the region refocusing matrix combined with the convolutional feature maps of the teacher network and the student network to perform scaling penalty training on the student network until the object detection accuracy of the student network converges. The marking module acquires an image of the user's hand within a fixed range near the integrated operation panel, inputs it into the student network, identifies the fingertips, and marks them to obtain marker points. The image acquisition module specifically includes: a first image acquisition module and a second image acquisition module. The first image acquisition module and the second image acquisition module respectively perform image acquisition of the marker points at a preset step size, and transmit the acquired images to the processing module. The processing module specifically includes: a coordinate acquisition module, a marker point coordinate system establishment module, and a region determination module; The coordinate acquisition module performs feature point matching on the first and second marked images acquired during the initial image acquisition by the first and second image acquisition modules to obtain the feature point coordinates in the coordinate system of the first image acquisition module. Coordinates of the corresponding feature points in the coordinate system of the second image acquisition module ; Calculate the unit direction vectors from the first and second image acquisition modules to the matching point in world coordinates. and : ; ; in, , These represent the coefficients used to normalize the direction vectors of the first and second image acquisition modules, respectively. , These represent the rotation matrices of the coordinate systems of the first image acquisition module and the second image acquisition module relative to the world coordinate system, respectively. , These represent the internal parameters of the first image acquisition module and the second image acquisition module, respectively, with superscript indicating their relative positions. Indicates transpose; Calculate the spatial coordinates of the first marked point : in, , , These represent the first image acquisition module. Second image acquisition module Spatial coordinates in the world coordinate system , They are rays and The foot of the perpendicular to the common perpendicular; The marker point coordinate system establishment module uses the spatial coordinates of the initial marker point. Establish the marker point coordinate system with the origin as the marker point coordinate system, and acquire the marker point image with the preset step size to obtain the spatial coordinates of the last acquired marker point. Obtain spatial vectors ; The region determination module divides the integrated operation panel into blocks according to the touch screen area corresponding to each icon, and inputs the coordinate system of the marked points, based on the spatial vector. The icon that passes through the touchscreen area is sent to the control module to turn on the corresponding range hood button.

2. The touch-sensitive range hood controller according to claim 1, characterized in that, It also includes a pollution detection module, which detects pollution in the data acquisition module, sets a pollution threshold, and issues an alarm when the pollution in the data acquisition module reaches the pollution threshold.

3. A touch-sensitive range hood controller according to claim 1, characterized in that, The integrated operation panel is provided with a contact area and a non-contact area. The contact area is provided with icons for corresponding buttons, and the non-contact area is provided with a data acquisition module.

Images