Projection device and anti-eye projection method

Abstract

The application provides a projection device and an anti-eye projection method. When the projection device acquires a target image, the method marks a pixel point corresponding to a character region in the target image as a target pixel point, generates a target circumscribed rectangle according to an image region formed by the target pixel point, acquires a minimum height coordinate and a maximum height coordinate according to a first top corner point and a second top corner point of the target circumscribed rectangle, calculates a head height of the character region according to an image height of the target image, a preset constant, the minimum height coordinate and the maximum height coordinate, determines a head region corresponding to the character region in the target circumscribed rectangle according to the head height, and reduces the brightness of projection content projected by a light-emitting assembly to the head region. The application classifies pixel points in a target image, identifies a head region of a character in a low-resolution target image, adjusts the brightness of projection content corresponding to the head region, and avoids strong light from irradiating a user's face.

Description

Technical Field

[0001] This application relates to the field of projection equipment technology, and in particular to a projection device and an anti-glare projection method. Background Technology

[0002] A projection device is a device that projects images, videos, and other information onto a screen or white wall using an optical system. It can be used for movie playback, video presentations, and image projection. Projection devices can project large-screen images from a considerable distance, allowing users to view a clearer picture.

[0003] When a projection device is working, it converts digital image information into light signals, which are then projected onto the projection area. When a user enters the projection area, the light emitted by the device shines directly into the user's eyes, potentially causing vision impairment. To address this, projection devices can be equipped with an eye-protection function. This function identifies the user's head area when they enter the projection area and reduces the brightness of the light source in that area, minimizing the impact of strong light on the user's vision.

[0004] Identifying a user's head area requires specific recognition algorithms. These algorithms need to acquire high-resolution images of the display area to identify the user's head within that image. However, low-end projection devices cannot acquire high-resolution images of the display area using high-resolution sensors. This prevents them from using specific recognition algorithms to determine the user's head area, thus hindering the application of anti-eye-spot protection features. Summary of the Invention

[0005] This application provides a projection device and an anti-glare projection method to solve the problem that low-resolution projection devices cannot apply the anti-glare function when a user enters the projection area.

[0006] In a first aspect, some embodiments of this application provide a projection device, including: The light-emitting component is configured to project the content onto the projection surface; A camera is configured to acquire a target image of the projection surface; The controller is configured as follows: When acquiring the target image, the pixels corresponding to the human figure area in the target image are marked as target pixels; Generate a target bounding rectangle based on the image region formed by the target pixels; Determine the first vertex point and the second vertex point opposite the first vertex point of the circumscribed rectangle, where the ordinate of the first vertex point is the minimum height coordinate and the ordinate of the second vertex point is the maximum height coordinate; The head height of the person's region is calculated based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates. The head height is the product of the height difference calculated from the maximum height coordinates and the minimum height coordinates and the preset constant. The head region corresponding to the person's region is determined within the target's bounding rectangle based on the head height. Reduce the brightness of the projected content onto the head area by the light-emitting component.

[0007] The above technical solution has the following beneficial effects or advantages: By performing pixel classification on the acquired target image, target pixels corresponding to the human figure region in the target image are separated, and a target bounding rectangle is generated in the image region formed by the target pixels. The height of the head region is determined by the target bounding rectangle. The projection device reduces the brightness of the projected content projected by the light-emitting component in the head region corresponding to the head height, thereby achieving the function of a projector eye. This embodiment, by recognizing the human head in a low-resolution target image, can reduce the projection brightness by recognizing the head region in low-resolution situations, thereby protecting the user's eyes from strong light when the user enters the projection area.

[0008] In some embodiments, the controller performs the step of marking the pixels corresponding to the person region in the target image as target pixels, specifically configured as follows: In response to a power-on command, the camera is controlled to acquire reference images; Obtain the reference pixel in the reference image; Based on the reference pixel, obtain an initial distance value representing the distance between the camera and the projection surface; If the difference between the initial distance value and the distance value of a pixel in the target image is greater than the determination threshold, then the pixel is marked as a target pixel.

[0009] The above technical solution has the following beneficial effects or advantages: When the projection device is powered on, a reference image is captured by the camera. When the user enters the projection area, if the difference between the initial distance value and the distance value of the pixel in the target image captured by the camera in real time is greater than the determination threshold, then the pixel is marked as a target pixel. Therefore, these pixels can be marked as target pixels, improving the efficiency of acquiring target pixels.

[0010] In some embodiments, the step of marking the pixel corresponding to the person region in the target image as the target pixel is specifically configured as follows: The target image is encoded by the encoder to obtain the target image code; The target image is encoded and input into the decoder corresponding to the encoder to obtain the image decoding result; Pixels whose binary classification results in the image decoding result are the same as the preset classification results are marked as target pixels. The binary classification result is either the probability result that the pixel in the image decoding result is a target pixel or the probability result that the pixel in the image decoding result is a non-target pixel.

[0011] The above technical solution has the following beneficial effects or advantages: by inputting the target image acquired in real time into the encoder for encoding, the pixel features in the target image are extracted to obtain the target image encoding, and the target image encoding is decoded by the corresponding decoder to obtain the image decoding result. Pixels in the image decoding result whose binary classification result is the same as the preset classification result are marked as target pixels, thereby improving the accuracy of obtaining target pixels.

[0012] In some embodiments, before the step of generating the target bounding rectangle based on the image region formed by the target pixels, the controller is further configured to: Obtain the region boundary points in the target pixel; The image region is obtained by sequentially connecting the adjacent boundary points of the region.

[0013] The above technical solution has the following beneficial effects or advantages: the projection device can obtain the regional boundary points in the target pixel points to determine the boundary of the person area in the projection area, and then connect the adjacent regional boundary points in sequence to obtain the image area of ​​the person entering the projection area.

[0014] In some embodiments, the step of obtaining the region boundary points in the target pixel is specifically configured as follows: Traverse the neighboring pixels of the target pixel; If the adjacent pixels include at least one non-target pixel, then the target pixel is marked as a region boundary point.

[0015] The above technical solution has the following beneficial effects or advantages: the region boundary point should be both the target pixel point and the non-target pixel point. To this end, the projection device can traverse the neighboring pixels of the target pixel point. If the neighboring pixels point includes at least one non-target pixel point, the target pixel point is marked as the region boundary point, and the image region is determined according to the region boundary point.

[0016] In some embodiments, the preset constant includes a target constant and a natural constant, and the step of calculating the head height of the person region based on the image height of the target image, the preset constant, the minimum height coordinates, and the maximum height coordinates is specifically configured as follows: Calculate the height difference between the maximum height coordinate and the minimum height coordinate; Calculate the height ratio of the height difference to the image height; The head height is obtained by calculating the product of the height difference, the target constant, and a hyperbolic tangent function with the height ratio as the exponent of the natural constant.

[0017] The above technical solution has the following beneficial effects or advantages: It calculates the height difference between the maximum and minimum height coordinates, so that the higher the person in the detected image, the closer the person is to the projection device, and the larger the proportion of the head should be. It then calculates the height ratio of the height difference to the image height, normalizing the height of the maximum bounding rectangle to 0-1. Finally, it calculates the head height based on the sum of the height differences, a preset constant, and the product of the hyperbolic tangent function.

[0018] In some embodiments, the step of determining the head region corresponding to the person region in the target bounding rectangle based on the head height is specifically configured as follows: Calculate the difference between the image height and the head height to determine the height of the area to be cropped; Based on the height of the area to be clipped, the target bounding rectangle is clipped to obtain the remaining rectangle after clipping; The target pixels in the remaining rectangle are marked as the head region.

[0019] The above technical solution has the following beneficial effects or advantages: by using the difference between the image height and the head height, the height of the area to be cropped is determined, and the torso area of ​​the person area is cropped according to the height of the area to be cropped, so as to preserve the head area of ​​the person area, thereby determining the projection area whose brightness needs to be reduced.

[0020] In some embodiments, the step of reducing the brightness of the projected content projected onto the head region by the light-emitting component is specifically configured as follows: Obtain the image dimensions of the head region in the target image; Based on the mapping relationship between the light-emitting component and the camera calibration, the image size is mapped to the area size of the light-emitting component to obtain the target processing area; Reduce the brightness of the projected content from the light-emitting component onto the target processing area.

[0021] The above technical solution has the following beneficial effects or advantages: In the above embodiment, the obtained head region is the image size. In order to map the image size to the projection size of the light-emitting component, the projection device can obtain the image size of the head region in the target image, and then map the image size to the resolution size of the camera according to the mapping relationship between the light-emitting component and the camera, thereby completing the conversion between the image size and the projection size and obtaining the target processing area.

[0022] In some embodiments, after the step of reducing the brightness of the projected content projected onto the head region by the light-emitting component, the controller is further configured to: Extract the regional feature points of the target processing area; The target processing region of the next target image is obtained by identifying the feature points of the region in the next target image of the current target image; Reduce the brightness of the projected content of the target processing area of ​​the next target image projected by the light-emitting component.

[0023] The above technical solution has the following beneficial effects or advantages: the target processing area is the head area of ​​the person. By obtaining the regional feature points of the head area, when the projection device acquires the next target image, it can determine whether the position of the head area has changed based on the identified regional feature points. Thus, based on the changed position of the head area, the target processing area of ​​the next target image is determined, and the brightness of the projected content projected onto the target processing area of ​​the next target image by the light output component is reduced.

[0024] Secondly, some embodiments of this application also provide an anti-glare projection method, applied to the projection device described in the first aspect, the projection device including a light-emitting component configured to project projection content onto a projection surface; a camera configured to capture a target image of the projection surface; and a controller; the method includes: When acquiring the target image, the pixels corresponding to the human figure area in the target image are marked as target pixels; Generate a target bounding rectangle based on the image region formed by the target pixels; Determine the first vertex point and the second vertex point opposite the first vertex point of the circumscribed rectangle, where the ordinate of the first vertex point is the minimum height coordinate and the ordinate of the second vertex point is the maximum height coordinate; The head height of the person's region is calculated based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates. The head height is the product of the height difference calculated from the maximum height coordinates and the minimum height coordinates and the preset constant. The head region corresponding to the person's region is determined within the target's bounding rectangle based on the head height. Reduce the brightness of the projected content onto the head area by the light-emitting component.

[0025] As can be seen from the above technical solutions, this application provides a projection device and an anti-glare projection method. When the projection device acquires a target image, it marks the pixels corresponding to the person's area in the target image as target pixels. A target bounding rectangle is generated based on the image area formed by the target pixels. The minimum and maximum height coordinates are obtained based on the first and second vertices of the target bounding rectangle. The head height of the person's area is then calculated based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates. The head area corresponding to the person's area is determined within the target bounding rectangle based on the head height, thereby reducing the brightness of the projected content projected onto the head area by the light-emitting component. This application classifies pixels in the target image to identify the head area of ​​a person in a low-resolution target image, thereby adjusting the brightness of the projected content corresponding to the head area to avoid strong light illuminating the user's face. Attached Figure Description

[0026] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of the projection state of a projection device provided in some embodiments of this application; Figure 2 This is a schematic diagram of the projection device structure provided in some embodiments of this application; Figure 3 This is a schematic diagram of the optical engine architecture of a projection device provided in some embodiments of this application; Figure 4 This is a schematic diagram of the optical path of a projection device provided in some embodiments of this application; Figure 5 This is a schematic diagram of the system framework of a projection device provided in some embodiments of this application; Figure 6 A flowchart illustrating the anti-glare projection method performed by the projection device provided in some embodiments of this application; Figure 7 A timing diagram showing the anti-glare projection method performed by a projection device provided in some embodiments of this application; Figure 8 A schematic diagram illustrating the acquisition of target images by a projection device provided in some embodiments of this application; Figure 9 A flowchart illustrating a first embodiment of a projection device for marking target pixels, provided in some embodiments of this application; Figure 10 A flowchart illustrating a second embodiment of a projection device for marking target pixels provided in some embodiments of this application; Figure 11 A schematic diagram of the boundary points of the projection device marking area provided in some embodiments of this application; Figure 12 A schematic diagram illustrating the cropping of the head region rectangle by a projection device provided in some embodiments of this application; Figure 13 This is a schematic diagram illustrating how a projection device, provided in some embodiments of this application, reduces the projection brightness in the head area. Detailed Implementation

[0028] To make the objectives and implementation methods of this application clearer, the exemplary implementation methods of this application will be clearly and completely described below with reference to the accompanying drawings of the exemplary embodiments of this application. Obviously, the exemplary embodiments described are only some embodiments of this application, and not all embodiments.

[0029] It should be noted that the brief descriptions of terms in this application are only for the convenience of understanding the embodiments described below, and are not intended to limit the embodiments of this application. Unless otherwise stated, these terms should be understood in their ordinary and common meaning.

[0030] The terms "first," "second," "third," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar or related objects or entities, and do not necessarily imply a specific order or sequence, unless otherwise specified. It should be understood that such terms are interchangeable where appropriate.

[0031] The terms “comprising” and “having”, and any variations thereof, are intended to cover but not exclude inclusion, for example, a product or device that includes a range of components is not necessarily limited to all of the components that are clearly listed, but may include other components that are not clearly listed or that are inherent to such product or device.

[0032] The term "module" refers to any known or subsequently developed hardware, software, firmware, artificial intelligence, fuzzy logic, or combination of hardware and / or software code that is capable of performing the functions associated with that element.

[0033] The embodiments of this application can be applied to various types of projection devices. The following description will use a projector as an example to illustrate the projection device and the automatic focusing method.

[0034] A projector is a device that projects images or videos onto a screen. Projectors can connect to computers, cable TV networks, the internet, VCD (Video Compact Disc), DVD (Digital Versatile Disc Recordable), game consoles, DV camcorders, and other devices via various interfaces to play corresponding video signals. Projectors are widely used in homes, offices, schools, and entertainment venues.

[0035] Figure 1 A schematic diagram of the placement of a projection device according to an embodiment of this application is shown. Figure 2 A schematic diagram of the optical path of a projection device according to an embodiment of this application is shown.

[0036] In some embodiments, reference Figure 1 and Figure 2 This application provides a projection device including a projection screen 1 and a projection device 2. The projection screen 1 is fixed in a first position, and the projection device 2 is placed in a second position so that the image projected by the device matches the projection screen 1. The projection device includes a laser light source 100, an optical engine 200, a lens 300, and a projection surface 400. The laser light source 100 provides illumination for the optical engine 200, which modulates the light beam and outputs it to the lens 300 for imaging, projecting it onto the projection surface 400 to form a projected image. Since the laser light source 100, the optical engine 200, and the lens 300 are used together to emit projection light to project the image, in some embodiments of this application, the laser light source 100, the optical engine 200, and the lens 300 are collectively referred to as the light-emitting assembly.

[0037] In some embodiments, the laser source 100 of the projection device includes a laser assembly and an optical lens assembly. The light beam emitted by the laser assembly can pass through the optical lens assembly to provide illumination for the light-emitting assembly. For example, the optical lens assembly requires a high level of environmental cleanliness and airtightness; while the chamber in which the laser assembly is installed can be sealed with a lower level of dustproof sealing to reduce sealing costs.

[0038] In some embodiments, the light-emitting components of the projection device may include a blue light engine, a green light engine, and a red light engine, and may also include a heat dissipation system, a circuit control system, etc. It should be noted that in some embodiments, the light-emitting component of the projector may also be implemented using an LED light source.

[0039] Figure 3A schematic diagram of the circuit architecture of a projection device according to an embodiment of this application is shown. In some embodiments, the projection device may include a display control circuit 10, a laser light source 20, at least one laser driving component 30, and at least one brightness sensor 40. The laser light source 20 may include at least one laser corresponding to at least one laser driving component 30. Here, "at least one" refers to one or more, and "more than one" refers to two or more.

[0040] Based on this circuit architecture, the projection device can achieve adaptive adjustment. For example, by setting a brightness sensor 40 in the light output path of the laser light source 20, the brightness sensor 40 can detect the first brightness value of the laser light source and send the first brightness value to the display control circuit 10.

[0041] The display control circuit 10 can acquire the second brightness value corresponding to the driving current of each laser, and determine that the laser has a COD fault when the difference between the second brightness value and the first brightness value of the laser is greater than the difference threshold. Then the display control circuit can adjust the current control signal of the corresponding laser driving component until the difference is less than or equal to the difference threshold, thereby eliminating the COD fault of the blue laser. The projection device can eliminate the COD fault of the laser in a timely manner, reduce the damage rate of the laser, and improve the image display effect of the projection device.

[0042] Figure 4 A schematic diagram of the structure of a projection device according to an embodiment of this application is shown.

[0043] In some embodiments, the laser light source 20 in the projection device may include independently configured blue laser 201, red laser 202 and green laser 203. The projection device may also be called a three-color projection device. The blue laser 201, red laser 202 and green laser 203 are all modular lightweight (Mirai Console Loader, MCL) packaged lasers, which are small in size and facilitate compact arrangement of the optical path.

[0044] In some embodiments, the controller includes at least one of a central processing unit (CPU), a video processor, an audio processor, a graphics processing unit (GPU), RAM (Random Access Memory), ROM (Read-Only Memory), a first to an nth interface for input / output, a communication bus, etc.

[0045] In some embodiments, the projection device may be configured with a camera for working in conjunction with the projection device to adjust and control the projection process. For example, the camera configured with the projection device may be specifically implemented as a 3D camera or a binocular camera; when the camera is implemented as a binocular camera, it specifically includes a left camera and a right camera; the binocular camera can acquire the image and playback content presented on the screen corresponding to the projection device, i.e., the projection surface, which is projected by the light-emitting component built into the projection device.

[0046] When the projection device moves, its projection angle and distance to the projection surface change, which will cause the projected image to be distorted. The projected image will be displayed as a trapezoidal image or other distorted images. The projection device controller can achieve automatic trapezoidal correction based on the image captured by the camera by coupling the angle between the projection surfaces and the correct display of the projected image.

[0047] Figure 5 A schematic diagram of the system framework for display control of a projection device according to an embodiment of this application is shown.

[0048] In some embodiments, the projection device has the characteristics of a long-throw micro-projector, and its controller can control the display of the projected light image through a preset algorithm to achieve functions such as automatic keystone correction, automatic screen entry, automatic obstacle avoidance, automatic focus adjustment, and eye protection.

[0049] In some embodiments, the projection device is equipped with a gyroscope sensor; during the movement of the device, the gyroscope sensor can sense the position movement and actively collect displacement signals; then, through the system framework layer, the collected data is sent to the application service layer to support the application data required during user interface interaction and application interaction. The collected data can also be used by the controller for data calls in the algorithm service implementation.

[0050] In some embodiments, the projection device is equipped with a time-of-flight sensor. After the time-of-flight sensor collects the corresponding data, the data will be sent to the time-of-flight service corresponding to the service layer. After the time-of-flight service obtains the data, it will send the collected data to the application service layer through a process communication framework. The data will be used for data calls, user interfaces, program applications, and other interactive applications of the controller.

[0051] In some embodiments, the projection device is configured with a camera for acquiring images, which may be a binocular camera, a depth camera, or a 3D camera, etc.; the camera acquisition data is sent to a camera service, and then the camera service sends the acquired target image to a process communication framework and / or a projection device calibration service; the projection device calibration service can receive the camera acquisition data sent by the camera service, and the controller can call the corresponding control algorithm in the algorithm library for different functions to be implemented.

[0052] In some embodiments, data interaction is performed with the application service through a process communication framework, and the calculation results are then fed back to the correction service through the process communication framework. The correction service sends the obtained calculation results to the projection device operating system to generate control signaling, and sends the control signaling to the light output component control driver to control the operating conditions of the light output component and realize automatic correction of the displayed image.

[0053] In some embodiments, the projection device can project a large-screen image from a greater distance, allowing users to view a clearer picture. When operating, the projection device converts received digital image information into light signals. These light signals carry the image information to be displayed. The projection device can then project these light signals onto the projection area to display the received digital image, i.e., the projected image. The digital image can be a fixed single frame, or it can be a frame from a video medium. Different digital images are projected based on the real-time changes in the uploaded video frames.

[0054] The projection surface 400 can be the projection area in the above embodiments. The projection surface 400 can be a flat area such as a wall or screen that can display the projected image. Therefore, the projection surface 400 can serve as the projection area. Since the projection area is illuminated by the light source emitted by the projection device, when a user enters the projection area and faces the projection device, the light source will shine directly into the user's eyes, causing damage to the user's eyesight.

[0055] To enhance the safety of projection equipment for users' eyesight, projectors can be equipped with an anti-glare function. The projector can capture a real-time image of the target area using a camera. When a user enters the projection area, the projector can identify the head region of the person in the target image, thereby reducing the brightness of the light source in that area and minimizing the impact of strong light on the user's vision, thus protecting their eyesight.

[0056] Projection devices require specific recognition algorithms to identify a user's head region, such as machine learning algorithms, deep learning object detection or segmentation algorithms, or template matching algorithms. These algorithms need to be based on clear and distinct human head features in the target image. Therefore, the cameras in these projection devices equipped with anti-glare features include high-resolution image sensors capable of acquiring high-resolution target images to meet the requirements of the recognition algorithms to perform recognition based on clear head features.

[0057] For low-end projection devices, the cameras equipped with these devices are not equipped with high-resolution sensors. Therefore, the target images captured by the low-resolution cameras cannot clearly identify the head features of people. This means that low-end projection devices cannot determine the user's head area through specific recognition algorithms, and thus cannot apply the anti-eye-spot function.

[0058] To address the issue that low-resolution projection devices cannot apply anti-glare functionality when a user enters the projection area, some embodiments of this application provide a projection device comprising a light-emitting component, a camera, and a controller. The light-emitting component is configured to project projection content onto a projection surface 400, where the area corresponding to the projection surface 400 can be referred to as the projection area. The camera is configured to capture the target image projected onto the projection surface 200. Figure 6 A flowchart illustrating the anti-glare projection method performed by the projection device provided in this embodiment. Figure 7 This is a timing diagram showing the anti-glare projection method executed by the projection device provided in the embodiments of this application. The projection device follows the... Figure 7 The timing relationships shown apply to the anti-glare projection method. See also Figure 6 and Figure 7 The controller is configured to: S100: When acquiring the target image, the pixels corresponding to the human figure area in the target image are marked as target pixels.

[0059] To determine if a person has entered the projection area, the projection device can capture a target image of the projection surface using a camera after powering on. This target image is not the same as the projected image itself, but rather a snapshot of the projection area. When no user is in the projection area, the content of the target image should be identical to the projected content; in this case, the projection device does not need to use an eye-protection feature. Figure 8 As shown, when a user enters the projection area, the user should be positioned between the projection device and the projection surface 400, blocking the light beam emitted by the light component. In this case, in the target image acquired by the projection device, because a portion of the light beam is blocked by the user, the pixels corresponding to this blocked portion of the light beam will show a significant difference from the pixels projected onto the projection surface 400. (See also...) Figure 8 In the target image, each cell in the array represents a pixel. Pixels marked as 0 are the pixels corresponding to the projected image, while other pixels marked as 0 are used to represent pixels acquired after being occluded by the user.

[0060] To easily distinguish between the pixels of the target image projected onto the projection surface 400 and the pixels of the target image projected onto the user, this embodiment can mark the pixels corresponding to the person area in the target image as the target pixels. It is understood that the user can pre-adjust the interval between camera acquisition of the target image corresponding to the projection area in the projection device's settings menu. For example, the camera can acquire the target image every 1 second or every 5 seconds. In this way, the projection device can automatically control the camera to acquire the target image of the projection area according to the user-preset interval after power-on.

[0061] Each time a target image of the projection area is acquired, the controller can identify the pixels corresponding to the human figures in the target image and mark these pixels to obtain the target pixels. To improve the accuracy of target pixel identification, the projection device can identify them in at least two ways. The following describes two examples of target pixel identification methods.

[0062] Example 1: The projection device can respond to a power-on command and, upon powering on, control the camera to capture a reference image. This reference image is the image captured by the camera when no content is being projected. It is important to note that when capturing the reference image, there should be no objects obstructing the projection surface 400 within the projection area. To this end, the projection device can perform object detection on the projection area during reference image capture. If an object obstructing the projection surface 400 is detected, a prompt message can be generated to remind the user to move the object out of the projection area.

[0063] The reference image includes reference pixels. The projection device can obtain the initial distance value representing the distance between the camera and the projection surface 400 based on these reference pixels. That is, when there is no obstruction, the distance value of the pixel corresponding to the projected image projected by the projection device should be the initial distance value.

[0064] However, when a user enters the projection area, the projected image is obscured by the user, causing the distance value of the obscured pixels in the target image to be less than the initial distance value. Based on this scenario, the projection device can set a threshold for determining the target pixel. For example... Figure 9 As shown, the projection device can calculate the difference between the initial distance value and the distance value of a pixel in the target image each time it acquires the target image of the projection area. When the difference is greater than a judgment threshold, it indicates that a pixel in the target image is occluded by the user, causing the pixel's distance value to decrease. Consequently, the difference between the initial distance value and the distance value of the pixel in the target image increases, exceeding the judgment threshold. Therefore, pixels in the target image with a difference greater than the judgment threshold can be marked as target pixels.

[0065] In another implementation of Embodiment 1, the target pixel can also be determined by comparing the pixel values. It should be understood that when a pixel is obscured by the user, the pixel's displayed position is closer to the light-emitting component relative to the projection surface 400. Therefore, the pixel value increases due to the reduced projection distance. Based on the above scenario, the controller can also obtain the initial pixel values ​​of each pixel in the reference image and calculate the difference between the real-time acquired target image's pixel values ​​and the initial pixel values. If the difference is greater than a pixel value threshold, the pixel is marked as the target pixel.

[0066] Example 2: Projection devices can also identify target pixels in a target image using a convolutional neural network. To do this, the controller can input the target image into the convolutional neural network when acquiring it through a camera. The convolutional neural network can include an encoder and a decoder, where the encoder performs encoding on the target image to obtain the target image encoding. For example... Figure 10 As shown, the encoder can include a hierarchical structure formed by multiple convolution operation modules to perform encoding operations on the target image layer by layer. For example, when the encoder includes a first convolution operation module, a second convolution operation module, and a third convolution operation module, after the target image is input into the encoder, the first convolution operation module performs initial encoding on the target image to extract pixel features from the target image and outputs a first encoding result. The first encoding result output by the first convolution operation module is then input into the second convolution operation module to extract pixel features again and output a second encoding result. Finally, the second encoding result is input into the third convolution operation module to complete the encoding and output the encoded target image.

[0067] In some embodiments, the convolution operation module may sequentially include a convolution operation unit, a linear integration unit, a depth-separable convolution operation, a linear integration unit, and a convolution operation unit. When the convolution operation module encodes the target image, it may perform operations sequentially according to the above unit structure.

[0068] After encoding is complete, the output target image can be input into the decoder corresponding to the encoder to obtain the image decoding result. See also... Figure 10 The decoder includes a hierarchical structure formed by multiple transposed convolution operation modules. During the decoding process, the target image can be directly encoded and decoded through the transposed convolution operation module. Alternatively, during the encoding process, the encoded result output by the convolution operation module can be added to the input of the next convolution operation module through matrix addition, and then input to the transposed convolution operation module to perform decoding.

[0069] In some embodiments, the convolution operation module may sequentially include a transposed convolution operation unit, a linear integration unit, a depthwise separable convolution operation, a linear integration unit, and a convolution operation unit. When the transposed convolution operation module encodes the target image, it may perform operations sequentially according to the above unit structure.

[0070] For example, in a convolution operation performed by a projection device, the output image (output) can be an array matrix of i rows and j columns, and the input image (input) consists of an array matrix of m rows and n columns and a convolution kernel. Therefore, the output image can be represented by the following formula: ; For a linear integration unit, it represents all pixels in the intermediate operation matrix undergoing one non-linear calculation. The non-linear function for a linear integration unit is: ; Where x is the pixel value corresponding to the input pixel.

[0071] In some embodiments, other nonlinear functions, such as the sigmoid function, tanh function, etc., can be used to replace the linear integration unit.

[0072] In some embodiments, the convolution operation module is executed sequentially by convolution calculation, linear integration unit, depthwise separable convolution operation, linear integration unit, and convolution operation. The calculation process is regarded as the process of encoding or decoding the input image into abstract features.

[0073] In some embodiments, the convolutional neural network needs to predict the binary classification result of each pixel in the target image. Based on the above scenario, the convolutional neural network in this embodiment can be an untrained neural network or a neural network pre-trained using sample images of the same resolution, including both target and non-target pixels. For the neural network undergoing training, the training loss can be obtained by calculating cross-entropy, i.e.: ; In the above formula, 'a' represents the resolution of the sample image, and 'b' is used to determine the resolution of the label corresponding to the image result. The Sigmoid function is: .

[0074] In the cross-entropy calculation involving the sigmoid function, the final trained convolutional neural network model will result in the output i*j matrix values ​​falling between 0 and 1. A threshold can be set to transform the final output matrix into a distribution of 0s and 1s. For example, when the output is 1, the binary classification result is the first classification result, and the pixel is the target pixel; when the output is 0, the binary classification result is the second classification result, and the pixel is a non-target pixel. The binary classification result represents the probability that a pixel in the image decoding result is either a target pixel or a non-target pixel.

[0075] S200: Generate a target bounding rectangle based on the image region formed by the target pixels.

[0076] After identifying target pixels in the target image, the image area formed by these target pixels constitutes the person area. Therefore, the projection device needs to identify the head area within this person area to reduce the projection brightness of the head region. However, in scenarios using low-resolution cameras, the projection device cannot perform head region recognition using its algorithm. Therefore, after acquiring the target image, the projection device generates a target bounding rectangle within the graphic area formed by the target pixels, and this bounding rectangle includes all target pixels.

[0077] In some embodiments, the projection device needs to determine the region boundary points of the image region formed by the target pixels. To facilitate the determination of region boundary points, the projection device can mark the pixels in the target image as target pixels and non-target pixels, for example, as... Figure 10 As shown, the projection device can mark target pixels in the target image as 1 and non-target pixels as 0. Region boundary points are target pixels that form the boundary of the person's region. Therefore, region boundary points should be adjacent to both target and non-target pixels. To achieve this, the projection device can traverse the neighboring pixels of a target pixel. If all neighboring pixels of a target pixel are target pixels, it means that the target pixel is located at the center of the person's region and is not a target pixel forming the region boundary.

[0078] If the target pixel's neighboring pixels include at least one non-target pixel, then the target pixel is considered to form the region boundary. The projection device can then mark this target pixel as the region boundary point. (See [link to relevant documentation]). Figure 11 , Figure 11 In the diagram, the pixels in the shaded area are the boundary points of the region. The projection device can use the first boundary point obtained from the traversal as the starting point of the boundary. After obtaining the starting point, the projection device uses the starting point as the center point to traverse the adjacent pixels of the starting point, and uses the first adjacent pixel obtained from the traversal as the target pixel for the next detection. The detection process of detecting that the adjacent pixels contain at least one non-target pixel is repeated for this target pixel until all the boundary points of the target bounding rectangle are obtained.

[0079] After acquiring the region boundary points, the projection device can connect all the region boundary points. During the connection process, the projection device can determine the region boundary point that is closest to the region boundary point to be connected, and thus connect adjacent region boundary points sequentially. After the connection is completed, the region boundary points and other target pixels located within the region formed by the region boundary points together form the image region.

[0080] In some embodiments, the projection area may further include multiple image regions, meaning that at least two users enter the projection area. Taking two users entering the projection area as an example, the target image acquired by the projection device may include two closed image regions. Therefore, the projection device can generate target bounding rectangles based on the two different image regions, i.e., generate a first target bounding rectangle for the first image region and a second target bounding rectangle for the second image region. The region boundary points of the first image region and the second image region can be represented as: ; in, These are the boundary points of the first image region. These are the boundary points of the second image region.

[0081] S300: Determine the first vertex point of the target bounding rectangle and the second vertex point opposite the first vertex point.

[0082] In generating the target bounding rectangle, the projection device can traverse the maximum and minimum x-coordinates, maximum and minimum y-coordinates of all target pixels within the image area. Then, it determines the first vertex point based on the maximum and minimum y-coordinates, and the second vertex point based on the minimum and maximum y-coordinates. The second vertex point is the diagonal point opposite the first vertex point. The y-coordinate of the first vertex point is the minimum height coordinate, and the y-coordinate of the second vertex point is the maximum height coordinate. Thus, using the first and second vertex points as the two vertices of the target bounding rectangle, the target bounding rectangle is generated.

[0083] In some embodiments, the order in which the first vertex point and the second vertex point are determined can be reversed. For example, the first vertex point is determined based on the minimum x-coordinate and the maximum y-coordinate, and the second vertex point is determined based on the maximum x-coordinate and the minimum y-coordinate. In this case, the y-coordinate of the first vertex point is the maximum height coordinate, and the y-coordinate of the second vertex point is the minimum height coordinate.

[0084] This embodiment uses the ordinate of the first vertex point as the minimum height coordinate and the ordinate of the second vertex point as the maximum height coordinate as an example for illustrative purposes.

[0085] S400: Calculate the head height of the person's region based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates.

[0086] The head height is the product of the height difference calculated from the maximum and minimum height coordinates and a preset constant. During the head height calculation process, the image height of the target image and the preset constant can be obtained from the projection data. The preset constant can include a natural constant and a pre-defined constant, which is a fixed value, for example, 0.5. After obtaining the image height and the preset constant, combined with the minimum and maximum height coordinates, the head height in the target image can be calculated using the following formula: ; in, The height of the head region in the target image is K, which is the head height. K is a preset constant. The coordinates of the maximum height. Here, is the minimum height coordinate, e is the natural constant, and H is the image height.

[0087] In the above formula, the projection device can calculate the height difference between the maximum and minimum height coordinates, and the height ratio of the height difference to the image height, thereby normalizing the height of the target's bounding rectangle to 0-1. Then, the height ratio is used as input to the hyperbolic tangent function, i.e. Since the height of the target's bounding rectangle is normalized to 0-1, it can be ensured that when x∈[0,1], this function takes the value [0,0.76], forming a direct proportional function. In the hyperbolic tangent function, the height ratio is used as the exponent of the natural constant e. Finally, the head height is obtained by calculating the product of the height difference, the preset constant, and the hyperbolic tangent function with the height ratio as the exponent of the natural constant.

[0088] S500: Determine the head region corresponding to the person region in the target bounding rectangle based on the head height.

[0089] To reduce interference from image areas outside the head region when acquiring the head region, the projection device can perform cropping based on these areas. To do this, the projection device can calculate the difference between the image height and the head height to determine the height of the cropped area, which is the image area outside the head region.

[0090] After determining the height to be cut, such as Figure 12As shown, the target bounding rectangle can be cropped based on the area to be cropped, removing the rectangular areas outside the head region, such as the torso and leg regions, resulting in a remaining rectangle. The user's head region, which is part of the projection area, is located within this remaining rectangle. The projection device can iterate through the target pixels within the remaining rectangle. Since these target pixels form the human body region, after cropping the area outside the head region, the target pixels within the remaining rectangle are the same as those forming the head region. The projection device can then mark these target pixels within the remaining rectangle as the head region.

[0091] In some embodiments, to improve the accuracy of obtaining the head region, the projection device can also identify the boundary points of the head region within the remaining rectangle, that is, traverse the adjacent pixels of the target pixel in the remaining rectangle. When at least one adjacent pixel is not the target pixel, the target pixel is the region pixel of the head region. After determining the region pixel, the projection device can also generate a circumscribed rectangle for the region pixel, that is, the head circumscribed rectangle, thereby determining the head region based on the head circumscribed rectangle.

[0092] S600: Reduce the brightness of the projected content from the light-emitting component onto the head area.

[0093] like Figure 13 As shown, after determining the head region, the projection device can reduce the brightness of the projected content onto the head region by the light-emitting component, thus reducing the impact of strong light on the user's vision when entering the projection area. Since the size of the head region in the target image differs from the actual size projected by the projection device, the projection device needs to convert the head region into the area actually projected by the projection device after obtaining it. To this end, in some embodiments, the projection device can obtain the image size of the head region in the target image, then obtain the mapping relationship between the light-emitting component and the camera calibration. Based on this mapping relationship, the image size of the target image captured by the camera is mapped to the area size of the light-emitting component to obtain the target processing area. The target processing area is the projection area of ​​the head region in the target image captured by the camera, mapped to the corresponding position projected by the light-emitting component. The projection device thus reduces the brightness of the projected content onto the target processing area by the light-emitting component, minimizing the impact on the user's vision.

[0094] In some embodiments, if a user moves continuously within the projection area, their head area will also move accordingly. Therefore, if the projection device reduces the brightness of the projected content in the target processing area, strong light from other projection areas outside the current target processing area will still damage the user's vision when they move. To address this, after determining the target processing area, the projection device extracts regional feature points within that area to obtain the head features of the user entering the projection area. After determining the regional feature points, when the projection device acquires the next target image from the current target image using a camera, it identifies the regional feature points in the next target image to determine the position of the user's head area in the next target image, i.e., the target processing area. If the position of the target processing area in the next target image is the same as the position where the projection device first acquired the target processing area, it indicates that the user has not moved during the time interval between acquiring the target image; therefore, there is no need to change the position where the projection content brightness is reduced.

[0095] If the target processing area of ​​the next target image is different from the position of the target processing area when the projection device first acquired it, it means that the user has moved to another position during the time interval of acquiring the target image. Therefore, it is necessary to reduce the brightness of the projected content according to the user's movement trajectory, with the target processing area as the trajectory width, so as to follow the user's head area according to the user's movement path and protect the user's vision in real time when moving in the projection area.

[0096] In some embodiments, the projection device may also employ a target tracking algorithm to determine the user's position within the projection area. This target tracking algorithm refers to an algorithm that determines the head region based on the coordinates of the head region in a continuous target image. The target tracking algorithm can be an IoU tracking algorithm, or it can use a pre-trained Deep-SORT tracking algorithm, etc.

[0097] Some embodiments of this application also provide an anti-glare projection method applied to a projection device. To facilitate the execution of the method, the projection device should at least include a light-emitting component configured to project projection content onto a projection surface; a camera configured to acquire a target image of the projection surface; and a controller. The method includes: S100: When acquiring the target image, the pixels corresponding to the human figure area in the target image are marked as target pixels.

[0098] S200: Generate a target bounding rectangle based on the image region formed by the target pixels.

[0099] S300: Determine the first vertex point of the circumscribed rectangle and the second vertex point opposite the first vertex point, wherein the ordinate of the first vertex point is the minimum height coordinate and the ordinate of the second vertex point is the maximum height coordinate.

[0100] S400: Calculate the head height of the person's region based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates.

[0101] The head height is the product of the height difference calculated from the maximum and minimum height coordinates and a preset constant.

[0102] S500: Determine the head region corresponding to the person region in the target bounding rectangle based on the head height.

[0103] S600: Reduce the brightness of the projected content from the light-emitting component onto the head area.

[0104] This application provides a projection device and an anti-glare projection method. When the projection device acquires a target image, it marks the pixels corresponding to the human figure area in the target image as target pixels. A target bounding rectangle is generated based on the image area formed by the target pixels. The minimum and maximum height coordinates are obtained based on the first and second vertices of the target bounding rectangle. The head height of the human figure area is then calculated based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates. The head area corresponding to the human figure area is determined within the target bounding rectangle based on the head height, thereby reducing the brightness of the projected content projected onto the head area by the light-emitting component. This application classifies pixels in the target image to identify the head area of ​​a human figure in a low-resolution target image, thereby adjusting the brightness of the projected content corresponding to the head area to prevent strong light from illuminating the user's face.

[0105] Similar parts between the embodiments provided in this application can be referred to mutually. The specific implementation methods provided above are only a few examples under the overall concept of this application and do not constitute a limitation on the scope of protection of this application. For those skilled in the art, any other implementation methods extended from the solution of this application without creative effort shall fall within the scope of protection of this application.

Claims

1. A projection device, characterized in that, include: The light-emitting component is configured to project the content onto the projection surface; A camera is configured to acquire a target image of the projection surface; The controller is configured as follows: When acquiring the target image, the pixels corresponding to the human figure area in the target image are marked as target pixels; Generate a target bounding rectangle based on the image region formed by the target pixels; Determine the first vertex point and the second vertex point opposite the first vertex point of the bounding rectangle of the target, where the ordinate of the first vertex point is the minimum height coordinate and the ordinate of the second vertex point is the maximum height coordinate; The head height of the person's region is calculated based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates. The head height is the product of the height difference calculated from the maximum height coordinates and the minimum height coordinates and the preset constant. Based on the head height, the head region is determined within the figure region of the target bounding rectangle; Reduce the brightness of the projected content onto the head area by the light-emitting component; The preset constants include target constants and natural constants. Specifically, the controller is configured to perform the step of calculating the head height of the person's region based on the image height of the target image, the preset constants, the minimum height coordinates, and the maximum height coordinates. Calculate the height difference between the maximum height coordinate and the minimum height coordinate; Calculate the height ratio of the height difference to the image height; The head height is obtained by calculating the product of the height difference, the target constant, and a hyperbolic tangent function whose exponent is the height ratio.

2. The projection device according to claim 1, characterized in that, The controller performs the step of marking the pixels corresponding to the human figure region in the target image as target pixels, specifically configured as follows: In response to a power-on command, the camera is controlled to acquire reference images; Obtain the reference pixel in the reference image; Based on the reference pixel, obtain an initial distance value representing the distance between the camera and the projection surface; If the difference between the initial distance value and the distance value of a pixel in the target image is greater than the determination threshold, then the pixel is marked as a target pixel.

3. The projection device according to claim 1, characterized in that, The controller performs the step of marking the pixels corresponding to the human figure region in the target image as target pixels, specifically configured as follows: The target image is encoded by an encoder to obtain the target image code; The target image is encoded and input into the decoder corresponding to the encoder to obtain the image decoding result; Pixels whose binary classification results in the image decoding result are the same as the preset classification results are marked as target pixels. The binary classification result is either the probability result that the pixel in the image decoding result is a target pixel or the probability result that the pixel in the image decoding result is a non-target pixel.

4. The projection device according to claim 1, characterized in that, Before the step of generating the target bounding rectangle based on the image region formed by the target pixels, the controller is further configured to: Obtain the region boundary points in the target pixel; The image region is obtained by sequentially connecting the adjacent boundary points of the region.

5. The projection device according to claim 4, characterized in that, The controller is specifically configured to perform the step of acquiring the region boundary points in the target pixel as follows: Traverse the neighboring pixels of the target pixel; If the adjacent pixels include at least one non-target pixel, then the target pixel is marked as a region boundary point.

6. The projection device according to claim 1, characterized in that, The controller executes the step of determining the head region corresponding to the person's region within the target bounding rectangle based on the head height, specifically configured as follows: Calculate the difference between the image height and the head height to determine the height of the area to be cropped; Based on the height of the area to be clipped, the target bounding rectangle is clipped to obtain the remaining rectangle after clipping; The target pixels in the remaining rectangle are marked as the head region.

7. The projection device according to claim 1, characterized in that, The controller is specifically configured to perform the step of reducing the brightness of the projected content from the light-emitting component onto the head region as follows: Obtain the image dimensions of the head region in the target image; Based on the mapping relationship between the light-emitting component and the camera calibration, the image size is mapped to the area size of the light-emitting component to obtain the target processing area; Reduce the brightness of the projected content from the light-emitting component onto the target processing area.

8. The projection device according to claim 7, characterized in that, After the controller performs the step of reducing the brightness of the projected content projected onto the head region by the light-emitting component, it is further configured to: Extract the regional feature points of the target processing area; The target processing region of the next target image is obtained by identifying the feature points of the region in the next target image of the current target image; Reduce the brightness of the projected content of the target processing area of ​​the next target image projected by the light-emitting component.

9. A method for preventing eye-spotting projection, characterized in that, The projection device according to any one of claims 1-8, the projection device including a light-emitting component, is configured to project projection content onto a projection surface; A camera is configured to acquire a target image of the projection surface; Controller; the method includes: When acquiring the target image, the pixels corresponding to the human figure area in the target image are marked as target pixels; Generate a target bounding rectangle based on the image region formed by the target pixels; Determine the first vertex point and the second vertex point opposite the first vertex point of the circumscribed rectangle, where the ordinate of the first vertex point is the minimum height coordinate and the ordinate of the second vertex point is the maximum height coordinate; The head height of the person's region is calculated based on the image height of the target image, a preset constant, the minimum height coordinates, and the maximum height coordinates. The head height is the product of the height difference calculated from the maximum height coordinates and the minimum height coordinates and the preset constant. Based on the head height, the head region is determined within the figure region of the target bounding rectangle; Reduce the brightness of the projected content onto the head area by the light-emitting component; The preset constants include target constants and natural constants. The controller executes the step of calculating the head height of the person's region based on the image height of the target image, the preset constants, the minimum height coordinates, and the maximum height coordinates, including: Calculate the height difference between the maximum height coordinate and the minimum height coordinate; Calculate the height ratio of the height difference to the image height; The head height is obtained by calculating the product of the height difference, the target constant, and a hyperbolic tangent function whose exponent is the height ratio.

Images