Screen light distribution calibration method and device, model training method and device, medium and product

Abstract

Embodiments of the present application provide a screen light distribution calibration method, a model training method, an apparatus, a device, a medium and a product. The method comprises: obtaining calibration images corresponding to a plurality of screens, inputting the calibration images into a region extraction model, and obtaining position information of the plurality of screens in the calibration images; and obtaining interval lighting images after each time of lighting a plurality of LEDs in the plurality of screens by lighting the plurality of LEDs each time. The method comprises: determining, by the region extraction model, image blocks corresponding to screen regions in each interval lighting image according to position information corresponding to the plurality of screens respectively, inputting the image blocks into a calibration model, and obtaining light distribution calibration data corresponding to the plurality of screens respectively. The method improves the light distribution calibration efficiency and improves the accuracy of the light distribution calibration data.

Description

Technical Field

[0001] This application relates to the field of optical inspection, and in particular to a screen light distribution calibration method, model training method, device, equipment, medium and product. Background Technology

[0002] Currently, when using LEDs (such as miniLEDs) as backlights for LCD screens, the LED backlight is divided into multiple independently controlled blocks so that the LEDs in each block can be independently dimmed according to the needs of the displayed content. The precise calibration of the LEDs is the main factor affecting the dimming effect.

[0003] In the prior art, the light distribution of each LED in the backlight of the LCD screen is obtained by individually lighting up each LED and calibrating the individual LEDs after they are lit.

[0004] However, due to the large number of LEDs in each screen, existing technologies require a long light distribution calibration time when calibrating LEDs on multiple screens, thus reducing the efficiency of light distribution calibration. Summary of the Invention

[0005] This application provides a screen light distribution calibration method, model training method, apparatus, device, medium, and product to improve the efficiency of light distribution calibration and the accuracy of light distribution calibration data.

[0006] In a first aspect, embodiments of this application provide a screen light distribution calibration method, including:

[0007] Obtain calibration images corresponding to multiple screens, extract the model from the input region of the calibration images, and obtain the position information of multiple screens in the calibration images;

[0008] By lighting up multiple LEDs on multiple screens each time, an interval lighting image is obtained after each lighting cycle;

[0009] The region extraction model determines the image blocks corresponding to the multiple screen regions in each spaced-lit image based on the location information of the multiple screens.

[0010] Multiple image patches are input into the calibration model to obtain light distribution calibration data corresponding to multiple screens.

[0011] Optionally, the calibration image input region extraction model is used to obtain the position information of multiple screens in the calibration image, including:

[0012] The calibration image is processed by a region extraction model to obtain the corner position information of each screen in the calibration image;

[0013] The position information of the corresponding screen is determined based on the corner position information of each screen.

[0014] Optionally, the calibration model includes an encoding module and a decoding module;

[0015] The encoding module is a multi-feature channel encoding module, and the decoding module is a multi-feature channel decoding module. The feature channels are determined based on the number of LEDs lit in each interval lit image.

[0016] The encoding module is used to perform multi-feature channel encoding processing on any image block to obtain the encoded feature data corresponding to the image block.

[0017] The decoding module is used to decode the encoded feature data to obtain the channel stitching data corresponding to the image block; the channel stitching data is the result of stitching together the light distribution calibration data of multiple currently lit LEDs in the image block in the channel dimension.

[0018] Optionally, the calibration model may also include dimension-reduced convolutional layers;

[0019] The dimensionality reduction convolutional layer is used to perform channel dimensionality reduction processing on the channel stitching data output by the decoding module to obtain the light distribution calibration data corresponding to each currently lit LED in the image block.

[0020] Optionally, after determining the image blocks corresponding to the multiple screen regions in each spaced-on image, the multiple image blocks are input into the corresponding calibration models, wherein each image block corresponds to a calibration model;

[0021] Multiple image blocks are processed using multiple calibration models to obtain light distribution calibration data for each LED in the multiple image blocks.

[0022] Optionally, multiple LEDs on multiple screens are lit at a time, including:

[0023] The number of LEDs corresponding to each of the multiple screens is determined based on the position information of each screen.

[0024] Based on the preset interval number of LEDs, determine the position information of multiple LEDs to be lit in each screen and light up multiple LEDs.

[0025] Secondly, embodiments of this application provide a model training method, including:

[0026] Obtain the training image set, which includes training calibration images corresponding to multiple training screens and multiple training interval illuminated images;

[0027] The calibration model is iteratively trained using a training image set and a loss function, which is determined based on the mask weight matrix.

[0028] The trained calibration model is used to obtain light distribution calibration data corresponding to multiple screens based on multiple image blocks. The multiple image blocks are determined based on multiple screen regions in each interval-lit image. The multiple screen regions are determined based on the position information corresponding to each screen. The interval-lit images are obtained by lighting multiple LEDs on multiple screens each time. The position information of the multiple screens is obtained by extracting the calibration image input region model corresponding to the multiple screens.

[0029] Optionally, the calibration model is iteratively trained using a training image set and a loss function, including:

[0030] By performing region extraction processing on the training calibration images, the position information corresponding to multiple training screens is obtained;

[0031] Based on the position information corresponding to multiple training screens, region extraction is performed on each training interval lit image to obtain image blocks corresponding to multiple training screens in each training interval lit image.

[0032] The calibration model is iteratively trained using multiple image patches corresponding to multiple training screens and loss functions.

[0033] The training interval lighting image is obtained by lighting up multiple LEDs on multiple training screens each time.

[0034] Optionally, before iteratively training the calibration model, the mask weight matrix of the LED is determined based on the preset effective area of ​​the light distribution of each LED in multiple training screens; the loss function used for iterative training is determined based on the mask weight matrix of each LED in multiple training screens.

[0035] Thirdly, embodiments of this application provide a screen light distribution calibration device, comprising:

[0036] The first acquisition model is used to acquire calibration images corresponding to multiple screens. The calibration image is input into the model to obtain the position information of multiple screens in the calibration image.

[0037] The first processing module is used to obtain the interval lighting image after each lighting by lighting multiple LEDs on multiple screens at a time;

[0038] The first processing module is also used to determine the image blocks corresponding to the multiple screen regions in each spaced-lit image based on the location information corresponding to the multiple screens using the region extraction model;

[0039] The first processing module is also used to input multiple image blocks into the calibration model to obtain light distribution calibration data corresponding to multiple screens.

[0040] Optionally, the first processing module is also used to perform recognition processing on the calibration image through a region extraction model to obtain the corner position information of each screen in the calibration image;

[0041] The position information of the corresponding screen is determined based on the corner position information of each screen.

[0042] Optionally, the calibration model includes an encoding module and a decoding module;

[0043] The encoding module is a multi-feature channel encoding module, and the decoding module is a multi-feature channel decoding module. The feature channels are determined based on the number of LEDs lit in each interval lit image.

[0044] The encoding module is used to perform multi-feature channel encoding processing on any image block to obtain the encoded feature data corresponding to the image block.

[0045] The decoding module is used to decode the encoded feature data to obtain the channel stitching data corresponding to the image block; the channel stitching data is the result of stitching together the light distribution calibration data of multiple currently lit LEDs in the image block in the channel dimension.

[0046] Optionally, the calibration model may also include dimension-reduced convolutional layers;

[0047] The dimensionality reduction convolutional layer is used to perform channel dimensionality reduction processing on the channel stitching data output by the decoding module to obtain the light distribution calibration data corresponding to each currently lit LED in the image block.

[0048] Optionally, the first processing module is further configured to, after determining the image blocks corresponding to the multiple screen areas in each interval lit image, input the multiple image blocks into the corresponding calibration models respectively, wherein each image block corresponds to a calibration model;

[0049] Multiple image blocks are processed using multiple calibration models to obtain light distribution calibration data for each LED in the multiple image blocks.

[0050] Optionally, the first processing module is further configured to determine the number of LEDs corresponding to each of the multiple screens based on the position information corresponding to each of the multiple screens.

[0051] Based on the preset interval number of LEDs, determine the position information of multiple LEDs to be lit in each screen and light up multiple LEDs.

[0052] Fourthly, embodiments of this application provide a model training apparatus, comprising:

[0053] The second acquisition module is used to acquire the training image set, which includes training calibration images corresponding to multiple training screens and multiple training interval lit images.

[0054] The second processing module is used to iteratively train the calibration model using the training image set and a loss function, which is determined based on the mask weight matrix.

[0055] The trained calibration model is used to obtain light distribution calibration data corresponding to multiple screens based on multiple image blocks. The multiple image blocks are determined based on multiple screen regions in each interval-lit image. The multiple screen regions are determined based on the position information corresponding to each screen. The interval-lit images are obtained by lighting multiple LEDs on multiple screens each time. The position information of the multiple screens is obtained by extracting the calibration image input region model corresponding to the multiple screens.

[0056] Optionally, the second processing module is further configured to obtain the position information corresponding to multiple training screens by performing region extraction processing on the training calibration image;

[0057] Based on the position information corresponding to multiple training screens, region extraction is performed on each training interval lit image to obtain image blocks corresponding to multiple training screens in each training interval lit image.

[0058] The calibration model is iteratively trained using multiple image patches corresponding to multiple training screens and loss functions.

[0059] The training interval lighting image is obtained by lighting up multiple LEDs on multiple training screens each time.

[0060] Optionally, the second processing module is further configured to, before iteratively training the calibration model, determine the mask weight matrix of the LEDs based on the preset effective light distribution area of ​​each LED in the multiple training screens; and determine the loss function used for iterative training based on the mask weight matrix of each LED in the multiple training screens. Fifthly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0061] The memory stores instructions that the computer executes;

[0062] The processor executes computer execution instructions stored in memory, causing the processor to perform various possible implementations of the first and / or second aspects described above.

[0063] In a sixth aspect, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement various possible implementations of the first and / or second aspects described above.

[0064] In a seventh aspect, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements various possible implementations of the first and / or second aspects described above.

[0065] The screen light distribution calibration method, model training method, apparatus, device, medium, and product provided in this application's embodiments acquire calibration images corresponding to multiple screens, input these images into a region extraction model, and obtain the position information of the multiple screens in the calibration images. By illuminating multiple LEDs in the multiple screens each time, an interval-lit image is obtained after each illumination. The region extraction model determines image blocks corresponding to the multiple screen regions in each interval-lit image based on the position information of each screen, and inputs these image blocks into the calibration model to obtain light distribution calibration data corresponding to each screen. This application improves the efficiency and accuracy of light distribution calibration data. Attached Figure Description

[0066] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0067] Figure 1 The prior art scenario diagram provided in this application;

[0068] Figure 2 Flowchart of the screen light distribution calibration method provided in this application Figure 1 ;

[0069] Figure 3 Flowchart of the screen light distribution calibration method provided in this application Figure 2 ;

[0070] Figure 4 A schematic diagram of the calibration model provided in this application;

[0071] Figure 5 Flowchart of the screen light distribution calibration method provided in this application Figure 3 ;

[0072] Figure 6 Flowchart of the model training method provided in this application Figure 1 ;

[0073] Figure 7 Flowchart of the model training method provided in this application Figure 2 ;

[0074] Figure 8 A schematic diagram of the screen light distribution calibration device provided in this application;

[0075] Figure 9 A schematic diagram of the structure of the model training device provided in this application;

[0076] Figure 10 A schematic diagram of the structure of the electronic device provided in this application.

[0077] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0078] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0079] First, let me explain the terms used in this application:

[0080] Light distribution: refers to the distribution of the emitted light from an LED on a screen.

[0081] Currently, the backlight of an LCD screen is composed of multiple LEDs (such as miniLEDs). When displaying images on an LCD screen, the backlight needs to be dimmed according to the display requirements of the image. Therefore, in order to improve the accuracy of image dimming, the light distribution of the LEDs in the screen backlight needs to be precisely calibrated.

[0082] Figure 1 The prior art scenario diagram provided in this application is as follows: Figure 1 As shown, the existing technology scenario includes a target screen, which comprises LED1, LED2, ..., LEDt, where t is a positive integer. By sequentially lighting up each LED in the target screen, a target LED image is obtained, and this target LED is the currently lit LED. By calibrating the target LED image, the light distribution calibration data of the currently lit target LED is obtained. Based on the above processing steps, the light distribution calibration data of each LED in the target screen is obtained. However, since each screen contains a large number of LEDs, the existing calibration method requires a significant amount of time when calibrating multiple screens, resulting in low light distribution calibration efficiency.

[0083] The screen light distribution calibration method provided in this application acquires calibration images corresponding to multiple screens and inputs these images into a region extraction model. The region extraction model identifies the corner positions of each screen in the calibration images, determines the position information of the corresponding screen based on the corner position information, and determines the image blocks corresponding to the multiple screen regions in each interval-lit image based on the position information of the multiple screens. The image blocks are then input into the calibration model, which performs calibration processing on each image block in each interval-lit image to obtain the light distribution calibration data corresponding to the multiple screens in each image block. This improves the efficiency of light distribution calibration. The calibration model is obtained after training a training model. During training, multiple training screens, training calibration images, and multiple training interval-lit images are input into the training model to decouple the multiple LEDs lit in the image block, reducing the mutual light emission influence between multiple LEDs lit in the same image block. This improves both the calibration efficiency and the accuracy and reliability of light distribution calibration for multiple screens.

[0084] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0085] Figure 2 Flowchart of the screen light distribution calibration method provided in this application Figure 1 ,like Figure 2 As shown, the method includes:

[0086] S201. Obtain calibration images corresponding to multiple screens, extract the model from the input region of the calibration images, and obtain the position information of multiple screens in the calibration images.

[0087] More specifically, by capturing images of each LED in multiple screens being lit, calibration images corresponding to multiple screens are obtained, and the calibration images are input into the model extraction area.

[0088] Optionally, the calibration image is processed by a region extraction model to obtain the corner position information of each screen in the calibration image; the position information of the corresponding screen is determined based on the corner position information of each screen.

[0089] For example, the corner position information is the coordinate information corresponding to the four corners of each screen.

[0090] In one possible embodiment, Figure 3 Flowchart of the screen light distribution calibration method provided in this application Figure 2 ,like Figure 3 As shown, a camera is placed in front of the moving track, and the shooting area on the moving track that the camera can capture is determined. Multiple screens currently to be calibrated (e.g., screen 1, screen 2, screen 3, and screen 4) are placed within this shooting area. Screens j, j+1, j+2, and j+3 on the moving track are the screens that have already been calibrated, while screens k, k+1, k+2, and k+3 are the screens awaiting calibration later.

[0091] In one possible embodiment, based on the respective positions of screens 1, 2, 3, and 4, each screen is determined to contain 30×30 LEDs. Here, 30×30 indicates that each screen has 30 LEDs distributed along both the horizontal and vertical axes. All LEDs in screens 1, 2, 3, and 4 are illuminated, and corresponding images are captured by a camera to obtain calibration images for screens 1, 2, 3, and 4. The calibration images are then processed using a region extraction model. Based on the illuminated LEDs in the calibration images, the corner position information of each screen is determined, thereby obtaining the position information of the corresponding screen. This embodiment uses a region extraction model to process the calibration images and determine the position information of each screen, achieving precise segmentation of the LED illumination images of multiple screens within the shooting area, further improving the accuracy and reliability of light distribution calibration for multiple screens.

[0092] For example, the corner position information includes, but is not limited to, the coordinates of the top left corner, the bottom left corner, the top right corner, and the bottom right corner, and the screen position information includes, but is not limited to, the screen edge information.

[0093] S202. By lighting up multiple LEDs on multiple screens each time, an interval lighting image is obtained after each lighting.

[0094] More specifically, based on the position information corresponding to each of the multiple screens, the number of LEDs corresponding to each of the multiple screens is determined; based on the preset interval number of LEDs, the position information of the multiple LEDs to be lit in each screen is determined and the multiple LEDs are lit, resulting in an interval lighting image.

[0095] In one possible embodiment, based on the respective positions of screens 1, 2, 3, and 4, each screen contains 30×30 LEDs, with a preset interval of 9 LEDs. Taking screen 1 as an example, the steps for the first interval lighting of screen 1 are as follows: Light up LED1 (the first one on the horizontal axis and the first one on the vertical axis); based on LED1, determine and light up LED2 (the eleventh one on the horizontal axis and the first one on the vertical axis), which is 9 LEDs apart on the horizontal axis; and determine and light up LED3 (the first one on the horizontal axis and the eleventh one on the vertical axis), which is 9 LEDs apart from LED1 on the vertical axis. Based on LED2 and LED3, LEDs spaced 9 LEDs apart from LED2 and LED3 along the horizontal axis are illuminated, and LEDs spaced 9 LEDs apart from LED2 and LED3 along the vertical axis are also illuminated, resulting in LED4 (21st LED on the horizontal axis, 1st LED on the vertical axis), LED5 (11th LED on the horizontal axis, 11th LED on the vertical axis), and LED6 (1st LED on the horizontal axis, 21st LED on the vertical axis). Based on LED4, LED5, and LED6, LEDs spaced 9 LEDs apart from LED4, LED5, and LED6 along the horizontal axis are illuminated, and LEDs spaced 9 LEDs apart from LED4, LED5, and LED6 along the vertical axis are also illuminated, resulting in LED7 (21st LED on the horizontal axis, 11th LED on the vertical axis), LED8 (21st LED on the horizontal axis, 11th LED on the vertical axis), and LED9 (21st LED on the horizontal axis, 21st LED on the vertical axis). This process is repeated to illuminate the LEDs on screens 2, 3, and 4 at intervals, resulting in the first interval-illuminated image.

[0096] For example, by analogy, multiple images of screens 1, 2, 3 and 4 lit at intervals are obtained within the shooting area until all LEDs in screens 1, 2, 3 and 4 have been lit, at which point the shooting of screens 1, 2, 3 and 4 is stopped.

[0097] Optionally, the shooting area and the number of screens that can be placed in the shooting area are determined according to the camera resolution and the size of the shooting field of view, and the number of screens that can be calibrated each time is determined based on the number of screens.

[0098] Optionally, such as Figure 3 As shown, when multiple screens to be calibrated are placed in the shooting area, the moving track is controlled to move. When the multiple screens to be calibrated are moved into the shooting area of ​​the camera, the moving track is controlled to stop moving and the multiple screens are kept stationary until the calibration process corresponding to the current multiple screens to be calibrated is completed. Then the moving track is restarted and the next group of multiple screens to be calibrated is transported to the shooting area and shot.

[0099] Optionally, the preset interval number in the horizontal and / or vertical axes is determined by calculating the quotient of the total number of LEDs on each screen in the horizontal and / or vertical axes and the number of LEDs that need to be lit on the screen each time. If the total number of LEDs on the screen in the horizontal and / or vertical axes is not divisible by the number of LEDs that need to be lit each time, the preset interval number is obtained by rounding up the calculated quotient.

[0100] In one possible embodiment, the total number of LEDs corresponding to the multiple screens within the shooting area can be the same or different.

[0101] In one possible embodiment, the LED lighting rules for multiple screens can be the same or different, wherein the lighting rules include, but are not limited to, a preset interval number of LEDs. For example, the preset interval number of LEDs for screen 1 is 9, the preset interval number of LEDs for screen 2 is 10, the preset interval number of LEDs for screen 1 is 11, and the preset interval number of LEDs for screen 2 is 12.

[0102] This embodiment captures multiple calibration images and multiple spaced-out illumination images of the screens within the camera's shooting area, enabling simultaneous calibration of multiple screens and illumination of each screen according to its corresponding illumination rules, thereby improving screen illumination efficiency and further enhancing light distribution calibration efficiency.

[0103] S203. Based on the location information corresponding to each screen in each spaced-lit image, the region extraction model is used to determine the image blocks corresponding to the multiple screen regions in each spaced-lit image.

[0104] More specifically, a region extraction model is used to segment each intermittently lit image based on the location information corresponding to multiple screens, resulting in image blocks corresponding to each screen region. The number of lit LEDs in multiple image blocks within the same intermittently lit image can be the same or different.

[0105] In one possible embodiment, such as Figure 3 As shown, the camera captures images of four screens within the shooting area to obtain calibration images and multiple intermittently lit images. The calibration images are then input into a region extraction model to determine the location information of each screen area. Based on the location information, intermittently lit images within the shooting area of ​​the four screens are obtained and processed to obtain four image blocks for each intermittently lit image, which are then input into the calibration model.

[0106] For example, taking the intermittent lighting image obtained from the first intermittent lighting in step S202 as an example, this intermittent lighting image includes 36 LEDs (9 LEDs are lit in screen 1, 9 LEDs are lit in screen 2, 9 LEDs are lit in screen 3, and 9 LEDs are lit in screen 4). The above intermittent lighting image is processed by a calibration model to obtain 4 image blocks. Taking image block 1 corresponding to screen 1 as an example, image block 1 is the image within the area of ​​screen 1 when 9 LEDs (LED1, LED2, LED3, LED4, LED5, LED6, LED7, LED8, and LED9 determined in step S201) are lit on screen 1.

[0107] S204. Input multiple image blocks into the calibration model to obtain light distribution calibration data corresponding to multiple screens.

[0108] More specifically, such as Figure 3 As shown, multiple image patches output by the region extraction model are input into the calibration model. The calibration model processes the image patches to obtain the light distribution calibration data corresponding to each screen and stores it.

[0109] Optionally, Figure 4 A schematic diagram of the calibration model provided in this application is shown below. Figure 4 As shown, the calibration model includes an encoding module, a decoding module, and a dimensionality reduction convolutional layer. The region extraction model processes the image data to be calibrated to obtain image blocks in each spaced-off illuminated image. The calibration model then processes each image block separately to obtain the light distribution calibration result corresponding to each image block. The light distribution calibration result for each image block includes the light distribution calibration data corresponding to the multiple currently illuminated LEDs within that image block.

[0110] Optionally, the encoding module is a multi-feature channel encoding module, and the decoding module is a multi-feature channel decoding module. The aforementioned feature channels are determined based on the number of LEDs lit in each interval-lit image. The encoding module is used to perform multi-feature channel encoding processing on any image block to obtain the encoded feature data corresponding to the image block; the decoding module is used to decode the encoded feature data to obtain the channel stitching data corresponding to the image block. The channel stitching data is the result of stitching together the light distribution calibration data of the multiple currently lit LEDs in the image block along the channel dimension. In one possible embodiment, the encoding module includes a 3x3 convolutional layer, a downsampling layer, and an activation function, and the decoding module includes an upsampling layer and a connection layer. The image block output by the region extraction model is input to the encoding module. The 3x3 convolutional layer of the encoding module extracts the local image features corresponding to the current image block and sends them to the downsampling layer. The downsampling layer performs downsampling processing on the local image features to obtain scale-invariant features, which are then sent to the activation function. The activation function enhances the nonlinearity of the current scale-invariant features to obtain nonlinearly enhanced multi-channel encoded feature data, which is then transmitted to the decoding module. The upsampling layer of the decoding module performs deconvolution processing on the encoded feature data to obtain multiple light distribution feature maps containing high-frequency detail information. These multiple light distribution feature maps are then stitched together by a connection layer according to the dimensions of the multiple feature channels to obtain the channel-stitched data corresponding to the aforementioned image patch. The high-frequency detail information includes the LED light distribution calibration data.

[0111] This embodiment uses a multi-feature channel encoding module and a decoding module to determine the light distribution feature maps corresponding to multiple LEDs in an image block, thereby obtaining the light distribution calibration data corresponding to multiple LEDs in the screen, which improves the light distribution calibration efficiency of the screen.

[0112] Optionally, the dimensionality reduction convolutional layer is used to perform channel dimensionality reduction processing on the channel stitching data output by the decoding module to obtain the light distribution calibration data corresponding to each currently lit LED in the image block.

[0113] In one possible embodiment, when there are nine LEDs lit at intervals in the current image block, and the dimensionality reduction convolutional layer is a 1x1 convolutional layer, the dimensionality reduction convolutional layer is used to perform dimensionality reduction processing on the stitched multiple light distribution feature maps, and outputs the light distribution calibration data corresponding to each of the nine LEDs. This embodiment obtains the light distribution calibration data corresponding to each lit LED in the current image block based on the dimensionality reduction convolutional layer, which improves the efficiency of light distribution calibration while improving the accuracy of light distribution calibration of the screen.

[0114] Optionally, after determining the image blocks corresponding to the multiple screen areas in each spaced-lit image, the multiple image blocks are input into the corresponding calibration models, where each image block corresponds to a calibration model; the multiple image blocks are processed by the multiple calibration models to obtain the light distribution calibration data corresponding to the multiple LEDs in the multiple image blocks.

[0115] In one possible embodiment, after obtaining image block 1, image block 2, image block 3, and image block 4 through a region extraction model, these four image blocks are sequentially input into the same calibration model. The calibration model then processes the image blocks accordingly, thereby obtaining the light distribution calibration data corresponding to the multiple LEDs lit on the screen for each image block.

[0116] In one possible embodiment, after obtaining image block 1, image block 2, image block 3, and image block 4 through a region extraction model, these four image blocks are simultaneously input into four calibration models with different parameters: image block 1 is input into calibration model 1, image block 2 into calibration model 2, image block 3 into calibration model 3, and image block 4 into calibration model 4. By simultaneously processing the image blocks through calibration models 1, 2, 3, and 4, light distribution calibration data corresponding to the multiple LEDs lit on the screen for each image block are obtained.

[0117] In one possible embodiment, after obtaining image block 1, image block 2, image block 3, and image block 4 through a region extraction model, these four image blocks are simultaneously input into four calibration models with identical parameters. The image blocks are processed through these four calibration models to simultaneously obtain the light distribution calibration data corresponding to the multiple LEDs lit on the screen for each image block.

[0118] This embodiment improves the efficiency of light distribution calibration on the screen.

[0119] The screen light distribution calibration method provided in this application acquires calibration images and intermittently lit images corresponding to multiple screens to be processed by capturing images, and inputs them into a region extraction model. The region extraction model determines the image block corresponding to each screen region in each intermittently lit image, and performs calibration processing using the image block as the calibration unit to obtain the light distribution calibration data of each screen, thereby improving the efficiency of screen light distribution calibration.

[0120] In one possible embodiment, Figure 5 Flowchart of the screen light distribution calibration method provided in this application Figure 3 ,like Figure 5As shown, by inputting the calibration images corresponding to multiple screens to be processed into the region extraction model, the region extraction model determines the position information corresponding to each screen. The interval-lit images corresponding to multiple screens are then input into the region extraction model, which determines the images of different screen regions in each interval-lit image based on the position information of each screen. This results in image blocks for each screen region under multiple interval-lit images, designated as image block 1, image block 2, image block 3, and image block 4. When image block 1 is used as the target image block, and image block 1 includes 9 lit LEDs, the target image block is input into the calibration model. The calibration model analyzes and processes the target image block to obtain the light distribution calibration data corresponding to each lit LED in the image block. Specifically, this yields calibration data 1 corresponding to the currently lit LED 1 in image block 1, calibration data 2 corresponding to the currently lit LED 2 in image block 1, ..., and calibration data 9 corresponding to the currently lit LED 9 in image block 1.

[0121] In one possible embodiment, when screen 1 and screen 2 exist within the shooting area, if a total of 100 images of the intermittently lit screens corresponding to screen 1 and screen 2 are captured, then the image block corresponding to screen 1 and the image block corresponding to screen 2 in each intermittently lit image are determined by a region extraction model. Taking screen 1 as an example, the light distribution calibration data corresponding to each lit LED in the image block corresponding to screen 1 in each intermittently lit image is output by a calibration model until the image block corresponding to screen 1 in the above 100 intermittently lit images is processed, thus obtaining the light distribution calibration data corresponding to screen 1.

[0122] This embodiment uses a region extraction model to determine each image block in each spaced-lit image after it has been divided into screen regions, and uses a calibration model to calibrate each image block to obtain the calibration data corresponding to the multiple LEDs currently lit in each image block. Therefore, it realizes the group calibration of screen LEDs, improves the efficiency of light distribution calibration, and ensures the accuracy of light distribution calibration.

[0123] Figure 6 Flowchart of the model training method provided in this application Figure 1 ,like Figure 6 As shown in this embodiment, the model training method is described in detail, and the method includes:

[0124] S601. Obtain the training image set.

[0125] More specifically, a training image set is obtained, which includes training calibration images corresponding to multiple training screens and multiple training interval illuminated images.

[0126] For example, the training calibration image is obtained by fully lighting up the LEDs of the training screen, and the training interval lighting image is obtained by lighting up multiple LEDs of multiple training screens at a time.

[0127] S602. The calibration model is iteratively trained using the training image set and loss function.

[0128] More specifically, the calibration model is iteratively trained using a training image set and a loss function, which is determined based on the mask weight matrix. The trained calibration model is used to obtain light distribution calibration data corresponding to multiple screens based on multiple image patches. The multiple image patches are determined based on multiple screen regions in each interval-lit image. The multiple screen regions are determined based on the position information corresponding to each screen. The interval-lit images are obtained by lighting multiple LEDs on multiple screens each time. The position information of the multiple screens is obtained by extracting the calibration image input region model corresponding to the multiple screens.

[0129] Optionally, by performing region extraction processing on the training calibration images, position information corresponding to multiple training screens is obtained; based on the position information corresponding to multiple training screens, region extraction is performed on each training interval lit image to obtain image blocks corresponding to multiple training screens in each training interval lit image; the calibration model is iteratively trained using multiple image blocks corresponding to multiple training screens and a loss function.

[0130] The model training method provided in this application calculates the corresponding mask weight matrix for each LED, updates the loss function of each training interval lit image based on the mask weight matrix, and iteratively trains the calibration model based on the updated loss function, thereby improving the calibration performance of the calibration model and further enhancing the accuracy and reliability of light distribution calibration data.

[0131] Optionally, before iteratively training the calibration model, the mask weight matrix of the LED is determined based on the preset effective area of ​​the light distribution of each LED in multiple training screens; the loss function used for iterative training is determined based on the mask weight matrix of each LED in multiple training screens.

[0132] For example, the loss function includes perceptual loss, absolute error loss, and relative error loss. For any image lit at intervals, the formula for calculating the loss function L is as follows:

[0133]

[0134] Where N represents the N LEDs lit in the interval-lit image, i represents the i-th LED lit in the interval-lit image, and L VGG For perceptual loss, L1 is the absolute error loss, and L...relative-L1 For L1-based relative error loss, mask i Let w1 be the mask weight matrix for the i-th LED, w2 be the weight coefficient used to characterize the perceptual loss, w3 be the weight coefficient used to characterize the absolute error loss, and w4 be the weight coefficient used to characterize the relative error loss.

[0135] For example, L relative-L1 The calculation formula is as follows:

[0136]

[0137] Where, x i f(x) represents the illuminated area corresponding to the i-th LED in the image, where f(x) is the area illuminated at intervals. i y represents the predicted light distribution calibration data for the i-th LED in the image lit at this interval. i The training light distribution calibration data for the i-th LED in the image lit at this interval.

[0138] Optionally, the steps for obtaining training light distribution calibration data are as follows: By lighting up each LED in the training screen one by one, the lighting image of each LED is obtained. Each lighting image is calibrated individually to obtain the training light distribution calibration data corresponding to that LED. This process is repeated to obtain the training light distribution calibration data corresponding to the training screen.

[0139] For example, the mask weight matrix includes multiple elements, with coordinates (a, b), and the value of each element is calculated using the following formula:

[0140]

[0141] Where S is the preset effective light distribution area for each LED.

[0142] In one possible embodiment, for a given LED, it is determined whether each element in the LED's mask weight matrix falls within a preset effective light distribution region of the LED. Elements whose coordinates fall within this region are assigned a value of 1; otherwise, the corresponding element is assigned a value of 0. By determining whether each element in the LED's mask weight matrix falls within the preset effective light distribution region, the values ​​of each element in the LED's mask weight matrix are obtained, and the loss function of the LED is calculated based on the LED's mask weight matrix.

[0143] In this embodiment, the loss function of the light distribution calibration task for each LED is updated by weighting the mask weight matrix corresponding to each LED. This allows for targeted adjustment of the calibration model's attention to different LED calibration tasks during training, thereby further enhancing the calibration performance of the calibration model.

[0144] Figure 7 Flowchart of the model training method provided in this application Figure 2 By lighting up each LED on the training screen one by one and calibrating each LED individually, training light distribution calibration data corresponding to each LED is obtained. Training calibration images are obtained by lighting up all LEDs on the training screen and taking pictures. Multiple training interval lighting images are obtained by lighting up multiple LEDs on the training screen at equal intervals in groups and taking pictures. The training image set is determined based on the multiple training interval lighting images and the training calibration images.

[0145] Optionally, the training image set is input into a region extraction model. The region extraction model processes the training calibration images to obtain the bounding box coordinates corresponding to each screen. Based on the bounding box coordinates, multiple image blocks corresponding to each training interval lit image are determined. By calibrating the distribution of these multiple image blocks in each training interval lit image, the predicted light distribution calibration data corresponding to the lit LEDs in each interval lit image is obtained. For example, ... Figure 7 As shown, for any interval-lit image, the calibration image includes 4 image blocks with a total of m lit LEDs. The calibration model calibrates each of the 4 image blocks in the interval-lit image. After the calibration of the 4 image blocks is completed, the model outputs the predicted light distribution calibration data 1, ..., predicted light distribution calibration data m corresponding to the interval-lit image. By inputting the predicted light distribution calibration data 1, ..., predicted light distribution calibration data m, and the training light distribution calibration data corresponding to the above m lit LEDs into the validation model, the validation model compares the predicted light distribution calibration data and training light distribution calibration data of each LED, determines the loss function according to the mask weight matrix corresponding to each of the above m lit LEDs, and updates the parameters of the calibration model iteratively according to the loss function. Here, m is a positive integer.

[0146] This embodiment updates the loss function by calculating the mask weight matrix corresponding to multiple LEDs in each training interval lit image, and iteratively trains the training model parameters for each training interval lit image and the corresponding loss function. This enhances the accuracy of the verification model in judging the authenticity of the above-mentioned predicted light distribution calibration data, as well as the authenticity of the predicted light distribution calibration data output by the calibration model, and further enhances the accuracy of the calibration model in calibrating the light distribution of the screen.

[0147] Figure 8 This is a schematic diagram of the screen light distribution calibration device provided in this application, as shown below. Figure 8 As shown, the screen light distribution calibration device 80 provided in this embodiment includes:

[0148] The first acquisition model 801 is used to acquire calibration images corresponding to multiple screens, extract the calibration image input region model, and obtain the position information of multiple screens in the calibration image;

[0149] The first processing module 802 is used to obtain the interval lighting image after each lighting by lighting multiple LEDs on multiple screens each time;

[0150] The first processing module 802 is also used to determine the image blocks corresponding to the multiple screen regions in each interval lit image based on the location information corresponding to the multiple screens using the region extraction model.

[0151] The first processing module 802 is also used to input multiple image blocks into the calibration model to obtain light distribution calibration data corresponding to multiple screens respectively.

[0152] Optionally, the first processing module 802 is further configured to perform recognition processing on the calibration image through a region extraction model to obtain the corner position information of each screen in the calibration image;

[0153] The position information of the corresponding screen is determined based on the corner position information of each screen.

[0154] Optionally, the calibration model includes an encoding module and a decoding module;

[0155] The encoding module is a multi-feature channel encoding module, and the decoding module is a multi-feature channel decoding module. The feature channels are determined based on the number of LEDs lit in each interval lit image.

[0156] The encoding module is used to perform multi-feature channel encoding processing on any image block to obtain the encoded feature data corresponding to the image block.

[0157] The decoding module is used to decode the encoded feature data to obtain the channel stitching data corresponding to the image block; the channel stitching data is the result of stitching together the light distribution calibration data of multiple currently lit LEDs in the image block in the channel dimension.

[0158] Optionally, the calibration model may also include dimension-reduced convolutional layers;

[0159] The dimensionality reduction convolutional layer is used to perform channel dimensionality reduction processing on the channel stitching data output by the decoding module to obtain the light distribution calibration data corresponding to each currently lit LED in the image block.

[0160] Optionally, the first processing module 802 is further configured to, after determining the image blocks corresponding to the multiple screen areas in each interval lit image, input the multiple image blocks into the corresponding calibration models respectively, wherein each image block corresponds to a calibration model;

[0161] Multiple image blocks are processed using multiple calibration models to obtain light distribution calibration data for each LED in the multiple image blocks.

[0162] Optionally, the first processing module 802 is further configured to determine the number of LEDs corresponding to each of the multiple screens based on the position information corresponding to each of the multiple screens.

[0163] Based on the preset interval number of LEDs, determine the position information of multiple LEDs to be lit in each screen and light up multiple LEDs.

[0164] The screen light distribution calibration device provided in this embodiment can execute the screen light distribution calibration method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0165] Figure 9 A schematic diagram of the structure of the model training device provided in this application is shown below. Figure 9 As shown, the model training device 90 provided in this embodiment includes:

[0166] The second acquisition module 901 is used to acquire a training image set, which includes training calibration images corresponding to multiple training screens and multiple training interval lit images.

[0167] The second processing module 902 is used to iteratively train the calibration model using the training image set and the loss function, which is determined based on the mask weight matrix.

[0168] The trained calibration model is used to obtain light distribution calibration data corresponding to multiple screens based on multiple image blocks. The multiple image blocks are determined based on multiple screen regions in each interval-lit image. The multiple screen regions are determined based on the position information corresponding to each screen. The interval-lit images are obtained by lighting multiple LEDs on multiple screens each time. The position information of the multiple screens is obtained by extracting the calibration image input region model corresponding to the multiple screens.

[0169] Optionally, the second processing module 902 is further configured to obtain the position information corresponding to multiple training screens by performing region extraction processing on the training calibration image;

[0170] Based on the position information corresponding to multiple training screens, region extraction is performed on each training interval lit image to obtain image blocks corresponding to multiple training screens in each training interval lit image.

[0171] The calibration model is iteratively trained using multiple image patches corresponding to multiple training screens and loss functions.

[0172] The training interval lighting image is obtained by lighting up multiple LEDs on multiple training screens each time.

[0173] Optionally, the second processing module 902 is further configured to determine the mask weight matrix of the LEDs based on the preset effective area of ​​the light distribution of each LED in the multiple training screens before iteratively training the calibration model; and to determine the loss function used for iterative training based on the mask weight matrix of each LED in the multiple training screens. The model training device provided in this embodiment can execute the model training method provided in the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described in detail here.

[0174] Figure 10 A schematic diagram of the structure of the electronic device provided in this application. Figure 10 As shown, the electronic device 100 provided in this embodiment includes at least one processor 1001 and a memory 1002. Optionally, the device 100 further includes a communication component 1003. The processor 1001, memory 1002, and communication component 1003 are connected via a bus 1004.

[0175] In a specific implementation, at least one processor 1001 executes computer execution instructions stored in memory 1002, causing at least one processor 1001 to perform the above-described method.

[0176] The specific implementation process of processor 1001 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0177] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0178] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0179] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0180] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0181] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0182] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0183] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0184] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0185] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0186] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0187] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0188] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0189] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for calibrating screen light distribution, characterized in that, include: Acquire calibration images corresponding to multiple screens, input the calibration images into a region extraction model, and obtain the position information of the multiple screens in the calibration images; By lighting up multiple LEDs on the multiple screens each time, an interval lighting image is obtained after each lighting. The region extraction model determines the image blocks corresponding to the multiple screen regions in each spaced-lit image based on the location information corresponding to the multiple screens. Multiple image blocks are input into the calibration model to obtain light distribution calibration data corresponding to the multiple screens respectively.

2. The method according to claim 1, characterized in that, The calibration image input region extraction model is used to obtain the position information of the multiple screens in the calibration image, including: The calibration image is processed by the region extraction model to obtain the corner position information of each screen in the calibration image; The position information of the corresponding screen is determined based on the corner position information of each screen.

3. The method according to claim 2, characterized in that, The calibration model includes an encoding module and a decoding module; The encoding module is a multi-feature channel encoding module, and the decoding module is a multi-feature channel decoding module. The feature channels are determined based on the number of LEDs lit in each interval lit image. The encoding module is used to perform multi-feature channel encoding processing on any image block to obtain the encoded feature data corresponding to the image block; The decoding module is used to decode the encoded feature data to obtain the channel stitching data corresponding to the image block; the channel stitching data is the result of stitching together the light distribution calibration data of multiple currently lit LEDs in the image block in the channel dimension.

4. The method according to claim 3, characterized in that, The calibration model also includes dimension-reduced convolutional layers; The dimensionality reduction convolutional layer is used to perform channel dimensionality reduction processing on the channel splicing data output by the decoding module to obtain the light distribution calibration data corresponding to each currently lit LED in the image block.

5. The method according to claim 1, characterized in that, Also includes: After determining the image blocks corresponding to multiple screen regions in each interval lit image, the multiple image blocks are input into the corresponding calibration models, wherein each image block corresponds to one calibration model; The multiple image blocks are processed by multiple calibration models to obtain light distribution calibration data corresponding to multiple LEDs in the multiple image blocks.

6. The method according to claim 2, characterized in that, Each time multiple LEDs on the multiple screens are lit, they include: The number of LEDs corresponding to each of the multiple screens is determined based on the position information corresponding to each of the multiple screens. Based on the preset interval number of LEDs, determine the position information of multiple LEDs to be lit in each screen and light up the multiple LEDs.

7. A model training method, characterized in that, include: Obtain a training image set, which includes training calibration images corresponding to multiple training screens and multiple training interval lit images; The calibration model is iteratively trained using the training image set and the loss function, wherein the loss function is determined based on the mask weight matrix; The trained calibration model is used to obtain light distribution calibration data corresponding to multiple screens based on multiple image blocks. The multiple image blocks are determined based on multiple screen regions in each interval-lit image. The multiple screen regions are determined based on the position information corresponding to each of the multiple screens. The interval-lit image is obtained by lighting multiple LEDs on the multiple screens each time. The position information of the multiple screens is obtained by extracting the calibration image input region model corresponding to the multiple screens.

8. The method according to claim 7, characterized in that, The calibration model is iteratively trained using the training image set and loss function, including: By performing region extraction processing on the training calibration image, the position information corresponding to multiple training screens is obtained; Based on the location information corresponding to the multiple training screens, region extraction is performed on each training interval lit image to obtain the image blocks corresponding to the multiple training screens in each training interval lit image. The calibration model is iteratively trained using multiple image patches corresponding to the multiple training screens and a loss function. The training interval lighting image is obtained each time multiple LEDs of the multiple training screens are lit.

9. The method according to claim 8, characterized in that, Also includes: Before iteratively training the calibration model, the mask weight matrix of the LED is determined based on the preset effective area of ​​the light distribution of each LED in multiple training screens. The loss function used in the iterative training is determined based on the mask weight matrix of each LED in the plurality of training screens.

10. A screen light distribution calibration device, characterized in that, include: The first acquisition module is used to acquire calibration images corresponding to multiple screens, input the calibration images into a region extraction model, and obtain the position information of the multiple screens in the calibration images; The first processing module is used to obtain an interval lighting image after each lighting by lighting multiple LEDs on the multiple screens each time; The first processing module is further configured to determine, based on the location information corresponding to the multiple screen regions in each interval lit image, the image blocks corresponding to the multiple screen regions in each image using the region extraction model; The first processing module is further configured to input multiple image blocks into the calibration model to obtain light distribution calibration data corresponding to the multiple screens respectively.

11. A model training device, characterized in that, include: The second acquisition module is used to acquire a training image set, which includes multiple training calibration images corresponding to multiple training screens and multiple training interval lit images. The second processing module is used to iteratively train the calibration model using the training image set; The trained calibration model is used to obtain light distribution calibration data corresponding to multiple screens based on multiple image blocks. The multiple image blocks are determined based on multiple screen regions in each interval-lit image. The multiple screen regions are determined based on the position information corresponding to each of the multiple screens. The interval-lit image is obtained by lighting multiple LEDs on the multiple screens each time. The position information of the multiple screens is obtained by extracting the calibration image input region model corresponding to the multiple screens.

12. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-9.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-9.

14. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method of any one of claims 1-9.

Images