Image content display optimization method and device, equipment and storage medium

Abstract

The invention belongs to the technical field of image processing, and relates to an image content display optimization method and device, equipment and a storage medium, and the method comprises the steps: carrying out the semantic segmentation of a preprocessed image, and obtaining a semantic segmentation region; performing thermodynamic diagram generation on the preprocessed image; fusing the semantic segmentation region and the image thermodynamic diagram to generate an arbitration weight map; converting the arbitration weight map into a to-be-displayed image; pre-scoring the to-be-displayed image according to a preset image quality scoring component to obtain a pre-scoring value; and when the pre-score value exceeds a preset score threshold value, displaying the to-be-displayed image to a target display interface. According to the method, after a target video is shot or the target video is synthesized, before the target video is displayed on a display interface in real time, in an image content optimization scene of each frame of real-time image in the target video before the target video is displayed, an arbitration weight map and an image quality score are introduced, so that the real-time image quality of the target video is improved. And the output of the displayed image with high image quality is ensured.

Description

Technical Field

[0001] This application relates to the field of image processing technology and is applied to the scenario of optimizing the display of real-time video stream image content using a visual processing model. It relates to an image content display optimization method, apparatus, device, and storage medium. Background Technology

[0002] Currently, mainstream display devices employ two main types of dynamic contrast enhancement technologies when enhancing images before and after display: one is global dynamic contrast enhancement, which relies on the global statistical characteristics of the image (such as histogram distribution and average pixel brightness) to dynamically generate a contrast adjustment curve applicable to the entire frame. This approach operates on a global scale and cannot resolve contrast optimization conflicts between different semantic objects within the same frame. The other type is fixed-zone local dynamic contrast enhancement, which pre-divides the image into fixed NxM physical grids (such as 16x9 or 32x18 squares) and independently performs statistical analysis and enhancement on each square to improve the shortcomings of the global approach. However, due to the significant segmentation, adjacent squares, with different processing parameters, are prone to brightness or color jumps at their boundaries, creating a visible "checkerboard" effect or halo, severely damaging image uniformity and immersion.

[0003] Therefore, in the existing technology, mainstream display devices still have problems with poor image quality adjustment when performing image enhancement before and after image display, and cannot guarantee the image output quality. Summary of the Invention

[0004] The purpose of this application is to provide an image content display optimization method, apparatus, device, and storage medium to solve the technical problem that mainstream display devices still have poor image quality adjustment and cannot guarantee image output quality when performing image enhancement before and after image display.

[0005] In a first aspect, embodiments of this application provide an image content display optimization method, which adopts the following technical solution: An image content display optimization method includes the following steps: Acquire the image to be processed and perform image preprocessing; Semantic segmentation is performed on the preprocessed image to obtain the semantic segmentation regions contained in the image; A heatmap is generated from the preprocessed image to obtain an image heatmap; By fusing the semantic segmentation region and the image heatmap, an arbitration weight map is generated; The arbitration weight graph is transformed and mapped to obtain the transformed image to be displayed; The image to be displayed is pre-scored according to a preset image quality scoring component to obtain a pre-scoring value; When the pre-score exceeds the preset score threshold, the image to be displayed is shown on the target display interface.

[0006] Secondly, embodiments of this application also provide an image content display optimization device, which adopts the following technical solution: An image content display optimization device includes: The image acquisition and preprocessing module is used to acquire the image to be processed and perform image preprocessing. The image semantic segmentation module is used to perform semantic segmentation on the preprocessed image to obtain the semantic segmentation regions contained in the image; An image heatmap generation module is used to generate a heatmap from the preprocessed image to obtain an image heatmap. The arbitration weight map generation module is used to fuse the semantic segmentation region and the image heatmap to generate an arbitration weight map; The image conversion module is used to convert and map the arbitration weight map to obtain the converted image to be displayed. The image quality scoring module is used to pre-score the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value; The image to be displayed module is used to display the image to be displayed on the target display interface when the pre-score value exceeds a preset score threshold.

[0007] Thirdly, embodiments of this application also provide a computer device that adopts the technical solution described below: A computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the image content display optimization method described above.

[0008] Fourthly, embodiments of this application also provide a computer-readable storage medium, which adopts the technical solutions described below: A computer-readable storage medium storing computer-readable instructions, which, when executed by a processor, implement the steps of the image content display optimization method described above.

[0009] Compared with the prior art, the embodiments of this application have the following main advantages: The image content display optimization method described in this application involves: acquiring an image to be processed and performing image preprocessing; performing semantic segmentation on the preprocessed image to obtain semantic segmentation regions contained in the image; generating a heatmap of the preprocessed image to obtain an image heatmap; fusing the semantic segmentation regions and the image heatmap to generate an arbitration weight map; performing a transformation mapping on the arbitration weight map to obtain a transformed image to be displayed; pre-scoring the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value; and displaying the image to be displayed on the target display interface when the pre-scoring value exceeds a preset scoring threshold. This method is used in scenarios where, after capturing or synthesizing a target video, but before displaying it on a real-time screen, image content optimization is performed on each frame of the target video before display. First, semantic segmentation is incorporated, improving upon the previous fixed segmentation shape to avoid a "chessboard" effect. Then, heatmaps are fused with semantic segmentation, and enhancement and suppression mechanisms are set for different weight regions after fusion to ensure that different semantic objects in the same image are transformed separately during image mapping. Finally, combined with a pre-processed image quality scoring method, the method ensures that the displayed image is output with high image quality by introducing an arbitration weight map and image quality scoring. Attached Figure Description

[0010] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is an exemplary system architecture diagram to which this application can be applied; Figure 2 This is a flowchart of an embodiment of an image content display optimization method according to this application; Figure 3 yes Figure 2 A flowchart of a specific embodiment of step 203 shown; Figure 4 yes Figure 2 A flowchart of a specific embodiment of step 205 shown; Figure 5 This is a schematic diagram of one embodiment of an image content display optimization device according to this application; Figure 6 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0012] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0013] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0014] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0015] like Figure 1 As shown, system architecture 100 may include terminal device 101, network 102, and server 103. Terminal device 101 may be a laptop 1011, tablet 1012, or mobile phone 1013. Network 102 is used as a medium to provide a communication link between terminal device 101 and server 103. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables.

[0016] Users can use terminal device 101 to interact with server 103 via network 102 to receive or send messages, etc. Various communication client applications can be installed on terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0017] Terminal device 101 can be various electronic devices with a display screen and support web browsing. In addition to laptops 1011, tablets 1012, or mobile phones 1013, terminal device 101 can also be an e-book reader, MP3 player (Moving Picture Experts Group Audio Layer III), MP4 player (Moving Picture Experts Group Audio Layer IV), laptop computer, and desktop computer, etc. The display screen is, for example, a TV display screen, a large display screen of a gaming device, laptop, or PC, or a curved display screen, etc. Through the image content display optimization method provided in this embodiment, real-time images can be mapped and converted to the target display screen with high-quality images.

[0018] Server 103 can be a server that provides various services, such as a backend server that provides support for the pages displayed on terminal device 101.

[0019] It should be noted that the image content display optimization method provided in this application embodiment is generally executed by a server, and correspondingly, the image content display optimization device is generally set in the server.

[0020] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0021] Continue to refer to Figure 2 The diagram illustrates a flowchart of an embodiment of an image content display optimization method according to this application. The image content display optimization method includes the following steps: Step 201: Obtain the image to be processed and perform image preprocessing.

[0022] In this embodiment, the image to be processed includes images from a video frame sequence or a continuous image stream that are to be displayed in real time.

[0023] Specifically, the image content display optimization method provided in this embodiment optimizes the image content of each frame of the target video in real time before displaying it on the display interface after the target video is captured or synthesized.

[0024] Specifically, the image preprocessing includes color space conversion of the image to be processed.

[0025] More specifically, the steps of acquiring the image to be processed and performing image preprocessing include: Real-time image extraction is performed on a video frame sequence or continuous image stream; the extracted real-time image is used as the image to be processed; and the color description of the image to be processed is converted from the initial color space to the target color space according to a preset color space conversion algorithm.

[0026] For example, the initial color space can be any one of RGB, YUV, or Ycbcr, with the specific initial color space determined by the video signal source. The target color space is the Ycbcr space. A color space is a way of describing color; different spaces emphasize different attributes. For example, the RGB space is based on the mixing of red, green, and blue light, while the Ycbcr space is closer to describing color saturation and brightness. For instance, in the Ycbcr space, Y represents the luminance component, and cb and cr represent the chrominance components. Performing this color space conversion allows for the conversion from RGB space to Ycbcr space, facilitating adjustments to image brightness and color saturation based on the luminance and chrominance components.

[0027] In this embodiment, by extracting the real-time image from a video frame sequence or a continuous image stream as the image to be processed, and performing color space conversion on the image to be processed, it is convenient to adjust the brightness and color saturation of the image using the luminance and chrominance components in the target color space.

[0028] Step 202: Perform semantic segmentation on the preprocessed image to obtain the semantic segmentation regions contained in the image.

[0029] In this embodiment, the preprocessed image, i.e. the image to be displayed in real time in a video frame sequence or continuous image stream, is the image after color space conversion.

[0030] Specifically, the semantic segmentation of the preprocessed image needs to consider semantic segmentation within the same image, i.e., the same image contains different objects, such as people, cars, and distant sky backgrounds. It also needs to consider the dynamic changes between consecutive frames and perform dynamic semantic segmentation. By combining the previous frame image, the dynamic changes generated from the previous frame image to the current real-time image can be identified. For example, from the previous frame image to the current real-time image, the arm of the person in the image has undergone dynamic changes.

[0031] More specifically, the semantic segmentation regions contained in the image include static segmentation regions corresponding to different objects in the same frame and dynamic change regions between consecutive frames. The step of performing semantic segmentation on the preprocessed image to obtain the semantic segmentation regions contained in the image specifically includes: using a preset lightweight convolutional neural network to perform static semantic segmentation on the preprocessed image to obtain the image regions occupied by different objects in the current real-time image; and using a preset lightweight convolutional neural network to perform dynamic semantic segmentation on the preprocessed image to obtain the image regions in the current real-time image that have dynamically changed compared to the previous frame. Finally, based on the image regions occupied by different objects in the current real-time image and the image regions in the current real-time image that have dynamically changed compared to the previous frame, corresponding semantic segmentation regions are divided in the image. This achieves the simultaneous use of static and dynamic semantic segmentation to divide the corresponding semantic segmentation regions in the current real-time image. The preset lightweight convolutional neural network refers to a CNN neural network with a relatively small number of layers. By introducing semantic segmentation, especially the combined processing of static and dynamic semantic segmentation, it is possible to divide the image regions occupied by different objects in the current real-time image, and also to segment the image regions that have dynamically changed compared to the previous frame. The combination of static and dynamic semantic segmentation avoids the "chessboard" noise caused by the previous "block segmentation".

[0032] In this embodiment, after performing dynamic semantic segmentation on the preprocessed image to obtain the semantic segmentation regions contained in the image, the method further includes: assigning a corresponding inertia attenuation factor to each semantic segmentation region according to the semantic segmentation regions contained in the image. Specifically, a smaller inertia attenuation factor is set for background regions in the semantic segmentation regions to ensure that the semantic segmentation regions respond quickly to changes in the scene; while a larger inertia attenuation factor is set for characters in the semantic segmentation regions to ensure that the semantic segmentation regions remain stable as the scene changes, that is, to ensure that they remain stable during the frame-by-frame changes of the video stream. For example, when the scene switches from an indoor drama to an outdoor landscape, an inertia factor α=0.95 is set for the character region parameters to allow for a slow transition in brightness and avoid a jarring effect; an α=0.5 is set for the background region parameters to allow for rapid adjustment and an immediate stunning effect.

[0033] Step 203: Generate a heatmap from the preprocessed image to obtain an image heatmap.

[0034] In this embodiment, the step of generating a heatmap from the preprocessed image to obtain an image heatmap is described below. Figure 3 , Figure 3 yes Figure 2A flowchart of a specific embodiment of step 203 shown includes: Step 301: Perform gridding processing on the preprocessed image to obtain a series of image grids; In this embodiment, the meshing process can be performed using... The preprocessed image is divided into a series of image grids. In the finest case, each image grid corresponds to only one pixel. Of course, the series of image grids can also be obtained directly according to pixel units.

[0035] Step 302: Using the visual saliency calculation method, calculate the attention value for each image grid, and statistically analyze the attention values ​​of all image grids to generate the attention value distribution. Specifically, the visual saliency calculation methods include the LC algorithm based on global contrast, the AC algorithm based on local contrast, or the FT algorithm based on the spectrum. The LC algorithm based on global contrast calculates the grayscale values ​​of all image grids using a grayscale image, and determines the visual saliency (i.e., the attention value) of all image grids based on the grayscale value calculation results. The AC algorithm based on local contrast calculates the local contrast values ​​of all image grids in the Lab color space, and determines the visual saliency (i.e., the attention value) of all image grids based on the local contrast values. The FT algorithm based on the spectrum analyzes the image in the frequency domain using Fourier transform, calculates the normalized kernel density estimate corresponding to all image grids, and determines the visual saliency (i.e., the attention value) of all image grids based on the normalized kernel density estimate.

[0036] In this embodiment, the LC algorithm based on global contrast can be directly used to calculate the gray values ​​of all image grids using grayscale image method. The attention values ​​of all image grids are determined by combining the gray value calculation results. The attention values ​​of all image grids are counted to generate the attention value distribution. Here, the larger the gray value of the corresponding image grid, the larger its attention value. Alternatively, the gray value can be directly set as the attention value.

[0037] Step 303: Generate thermal characterization values ​​for each of the image grids based on the attention value distribution. In this embodiment, the attention value can be directly proportional to the color intensity and brightness of the image, that is, the larger the attention value, the stronger the color intensity and the higher the brightness of the image; the thermal characterization value can also be directly the color intensity or brightness of the image.

[0038] Step 304: Generate the image heatmap based on the thermal characterization values ​​corresponding to each of the image grids.

[0039] Specifically, the grayscale values ​​of different image grids are directly set as their attention values. Based on the distribution of the attention values, image grids with the same or similar grayscale values ​​are identified. A heatmap is generated based on the attention values ​​corresponding to different image grids. The heatmap generation includes gradually increasing the brightness level or color brightness level of different image grids from low to high based on the grayscale values ​​corresponding to different image grids, thereby generating the image heatmap.

[0040] In this embodiment, an image heatmap is generated through visual saliency calculation, so that the generation of the image heatmap is in the mode of human visual saliency calculation, which satisfies human visual perception.

[0041] Step 204: Merge the semantic segmentation region and the image heatmap to generate an arbitration weight map.

[0042] Specifically, the arbitration weight map, based on the fusion of the semantic segmentation region and the image heatmap, introduces enhancement and suppression mechanisms for different semantic segmentation regions. This ensures that image mapping is performed using the appropriate mechanisms during subsequent display image mapping. Specifically, when certain semantic segmentation regions experience significant color intensity enhancement during image transformation mapping, potentially causing white light effects, a suppression mechanism is used to suppress the color intensity enhancement during transformation. Conversely, for some darker display areas, to avoid excessive differences in image color intensity within the same image, an enhancement mechanism can be used to enhance the color intensity of these display areas.

[0043] In this embodiment, the step of fusing the semantic segmentation region and the image heatmap to generate the arbitration weight map specifically includes: counting the image grids contained in each semantic segmentation region; and using a preset fusion mode, fusing the semantic segmentation region and the image heatmap to generate the arbitration weight map based on the image grids contained in each semantic segmentation region and the heatmap values ​​corresponding to all image grids.

[0044] In this embodiment, the semantic segmentation region and the image heatmap are fused. In the region division of the same image, the segmentation method is improved from fixed shape-based segmentation to semantic classification-based segmentation. At the same time, considering the intensity difference between different semantic segmentation regions in subsequent display mapping, in order to avoid the imbalance of image quality in the same display image, such as excessive white light or excessive darkness, the arbitration weight map is generated. It can perform interference enhancement or suppression adjustment on the mapping result during actual display mapping to ensure the quality of the display image.

[0045] Specifically, the step of using a preset fusion mode to fuse the semantic segmentation region and the image heatmap to generate the arbitration weight map includes: using the preset fusion mode to set the region weights of all semantic segmentation regions to obtain a region weight map; dividing the region weight map into high-weight regions and low-weight regions according to a preset region weight threshold; setting a preset inhibitory adjustment mechanism associated with all high-weight regions and a preset enhancement adjustment mechanism associated with all low-weight regions to obtain the arbitration weight map.

[0046] It should be understood that the same image may include multiple semantic categories, i.e., multiple semantic segmentation regions, such as people, vehicles, trees, and background regions. Within these semantic segmentation regions, trees and background regions are generally static objects in the video and do not undergo significant dynamic changes, while people and vehicles are generally dynamic objects in the video and may undergo significant dynamic changes. For example, a person walking involves changes in limb movement, and a vehicle moving also involves dynamic changes. When watching a video, the human eye tends to observe dynamic objects more. Therefore, based on a preset region weight threshold, high-weight regions and low-weight regions are divided in the region weight map. That is, semantic segmentation regions that the human eye pays more attention to are set as high-weight regions, and semantic segmentation regions that the human eye pays less attention to are set as low-weight regions. Furthermore, a preset inhibitory adjustment mechanism is set for all high-weight regions, and a preset enhancement adjustment mechanism is set for all low-weight regions. That is, a preset inhibitory adjustment mechanism is set for semantic segmentation regions that the human eye pays more attention to avoid excessive image enhancement of these semantic segmentation regions, while a preset enhancement adjustment mechanism is set for semantic segmentation regions that the human eye pays less attention to avoid insufficient image enhancement.

[0047] Specifically, for semantic segmentation areas that the human eye pays more attention to, such as brighter or darker areas, a preset suppression adjustment mechanism is invoked when displaying these semantic segmentation areas to prevent excessive enhancement and unbalanced screen brightness. For semantic segmentation areas that the human eye pays less attention to, such as background areas or normally displayed areas, a preset enhancement adjustment mechanism can be invoked when displaying these semantic segmentation areas to make their screen brightness closer to the areas of focus, thus avoiding polarization in the overall screen mapping.

[0048] More specifically, the preset fusion mode is used to set the regional weights of all semantic segmentation regions to obtain a regional weight map, including two fusion modes: one is a fusion mode based on the average value algorithm, and the other is a fusion mode based on the mode algorithm.

[0049] The fusion mode based on the average value algorithm is as follows: according to the image grids contained in each semantic segmentation region and the thermal characterization values ​​corresponding to all image grids, the average value algorithm is used to calculate the average thermal characterization value of the image grids in each semantic segmentation region, so as to obtain the average thermal characterization value corresponding to each semantic segmentation region; the average thermal characterization value corresponding to each semantic segmentation region is set as the region weight of the corresponding semantic segmentation region, so as to obtain the region weight map.

[0050] The fusion mode based on the modal value algorithm is as follows: according to the image grids contained in each semantic segmentation region and the thermal representation values ​​corresponding to all image grids, the modal value algorithm is used to calculate the modal thermal representation value of the image grids in each semantic segmentation region, so as to obtain the modal thermal representation value corresponding to each semantic segmentation region; the modal thermal representation value corresponding to each semantic segmentation region is set as the regional weight of the corresponding semantic segmentation region, so as to obtain the regional weight map.

[0051] In this embodiment, the modal algorithm, for example: if the face semantic segmentation region corresponds to 1000 pixels and the thermal representation value of 700 pixels is A, then the A value is directly set as the region weight of the face semantic segmentation region.

[0052] In this embodiment, the fusion mode can be selected according to the actual scenario requirements, and no limitation is made here.

[0053] Step 205: Transform and map the arbitration weight graph to obtain the transformed image to be displayed.

[0054] In this embodiment, the step of transforming and mapping the arbitration weight map to obtain the transformed image to be displayed is described in the following reference. Figure 4 , Figure 4 yes Figure 2 A flowchart of a specific embodiment of step 205 shown includes: Step 401: Use the region weight of each semantic segmentation region as the input parameter of the baseline gamma curve, and use the adjustment mechanism in the arbitration weight map as the adaptive interpolation compensation mechanism to construct a gamma mapping curve for the image to be displayed to be converted and mapped by the arbitration weight map. The adjustment mechanism in the arbitration weight map includes the preset inhibitory adjustment mechanism and the preset enhancement adjustment mechanism. Specifically, the baseline gamma curve conforms to the formula: The curve, in the formula, This represents the color intensity of each pixel in the image before it is displayed on the interface. This indicates the color intensity of the corresponding pixel when the image is displayed on the interface. For the transformation coefficients, here, the region weight of each semantic segmentation region, that is, the overall color intensity of each semantic segmentation region, is used as the input parameter to obtain the overall color output intensity of the corresponding semantic segmentation region.

[0055] To ensure relative controllability of color output intensity, the adjustment mechanisms in the arbitration weight map, namely the preset inhibitory adjustment mechanism for high-weight semantic segmentation regions and the preset enhancement adjustment mechanism for low-weight semantic segmentation regions, are introduced into the image transformation mapping.

[0056] Accordingly, the reference gamma curve is improved as follows: Here, 'a' is the enhancement adjustment value given by the preset enhancement adjustment mechanism, and 'b' is the suppression adjustment value given by the preset suppression adjustment mechanism. It can also be understood that 'a' makes the converted color intensity continue to increase, while 'b' makes the converted color intensity decrease. The color intensity includes luminance intensity and / or color saturation intensity. Accordingly, the values ​​of 'a' and 'b' can be adjusted according to actual details, and can be luminance component values ​​or color component values, respectively.

[0057] Step 402: The gamma mapping curve is continuously optimized using a continuous optimization method and the adaptive interpolation compensation mechanism until the preset optimization termination condition is met, at which point the continuous optimization stops and the final gamma curve is obtained. In this embodiment, the step of continuously optimizing the gamma mapping curve using a continuous optimization method and an adaptive interpolation compensation mechanism until a preset optimization termination condition is met, thereby stopping the continuous optimization and obtaining the final gamma curve, includes: detecting the intensity values ​​of all image grids before and after image transformation mapping based on the current gamma mapping curve using a preset intensity detection component; calculating the intensity change rate of each image grid after image transformation mapping based on the intensity values; if the intensity change rate is in an intensity unchanged state, no interpolation compensation is performed on the current image grid; if the intensity change rate is in an intensity-increasing state, the preset inhibitory adjustment mechanism is used to perform reverse interpolation compensation on high-weight semantic segmentation regions; if the intensity change rate is in an intensity-decreasing state, the preset enhancement adjustment mechanism is used to perform forward interpolation suppression on low-weight semantic segmentation regions; until all image grids no longer require interpolation compensation in terms of intensity change rate, the optimized and adjusted gamma curve is obtained as the final gamma curve.

[0058] Specifically, the gamma mapping curve is continuously optimized until it meets the preset image display requirements, such as: the color brightness of all semantic segmentation regions is in a transition state with good image quality and suitable color brightness, then the final gamma curve is obtained.

[0059] More specifically, the preset intensity detection components include, for example, color intensity detection components and image brightness intensity detection components; and based on the intensity values, the intensity change rate of each image grid after image transformation and mapping is calculated, using an intensity change rate conversion algorithm, for example: ,in, Indicates the output color intensity. Indicates the input color intensity. To be a minimum value, avoid A value of 0 results in a zero-value anomaly. Generally, if... If the value is 1, it indicates that the intensity change rate is in a state where the intensity has not changed, and therefore no interpolation compensation is performed on the current image grid. If the value is greater than 1, it indicates that the intensity change rate is in an intensity-increasing state, and the preset inhibitory adjustment mechanism is used to perform reverse interpolation compensation on the high-weight semantic segmentation region; if If the value is less than 1, it indicates that the intensity change rate is in a state of decreasing intensity. In this case, the preset enhancement adjustment mechanism is used to perform positive interpolation suppression on the low-weight semantic segmentation region.

[0060] In this embodiment, adaptive interpolation compensation for color intensity can be achieved by using adaptive interpolation compensation for chromaticity components in the Ycbcr space.

[0061] Step 403: Based on the final gamma curve, perform image transformation mapping on the arbitration weight map to obtain the transformed image to be displayed.

[0062] Specifically, the final gamma curve is used to perform a transformation mapping on the arbitration weight map to obtain the transformed image to be displayed.

[0063] During image display mapping, the curves for faces and subtitles are nearly linear, preserving more details; the curves for natural scenery are moderately steep, enhancing the sense of depth in the image; and the background curve is the steepest, revealing details. Using uniform interpolation parameters would be inapplicable. Therefore, this embodiment introduces a combination of adjustment and adaptive interpolation compensation mechanisms based on the fundamental gamma curve during image mapping. This adaptive interpolation process ensures smooth and seamless transitions between faces and backgrounds, and between text and scenery, completely avoiding the blockiness that may occur in fixed partitioning schemes.

[0064] Step 206: Perform a pre-scoring of the image to be displayed according to the preset image quality scoring component to obtain a pre-scoring value.

[0065] In this embodiment, the step of pre-scoring the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value specifically includes: obtaining the image quality parameter values ​​bound to each semantic segmentation region in the image to be displayed; and performing a comprehensive score on the image to be displayed based on the image quality parameter values ​​bound to each semantic segmentation region to obtain the pre-scoring value.

[0066] In this embodiment, the preset image quality scoring component binds image quality parameter values ​​for each semantic segmentation region during setup. After obtaining the image quality parameter values ​​bound to each semantic segmentation region, a comprehensive image quality score result can be obtained by summing or averaging the image quality parameter values ​​bound to each semantic segmentation region during calculation.

[0067] Specifically, image quality parameter values ​​are bound to each semantic segmentation region. Subsequently, when scoring the image conversion quality, it is only necessary to obtain the image quality parameter values ​​bound to all semantic segmentation regions to calculate the comprehensive score.

[0068] In this embodiment, the image quality parameter values ​​bound to each of the semantic segmentation regions change with the increase of the frame value following a first-order infinite impulse response filter trend, wherein the trend curve of the first-order infinite impulse response filter is as follows: ,in, It is the image quality parameter value after the corresponding semantic segmentation region in the previous frame has been smoothed out. It is the inertia attenuation factor of the corresponding semantic segmentation region. This is the reference image quality parameter value for the corresponding semantic segmentation region in the current frame image. This value can be set based on the thermal characterization value of the semantic segmentation region before image conversion. It is the final image quality parameter value of the corresponding semantic segmentation region in the current frame image.

[0069] In this embodiment, after performing the step of pre-scoring the image to be displayed according to the preset image quality scoring component to obtain the pre-scoring value, the method further includes: when the pre-scoring value does not exceed the preset scoring threshold, triggering a preset parameter rollback or fine-tuning mechanism to perform feedforward optimization until the pre-scoring value exceeds the preset scoring threshold and the feedforward optimization is completed.

[0070] Specifically, the mechanism for triggering preset parameter rollback or fine-tuning to perform feedforward optimization includes: selectively reprocessing the image to be processed, semantically segmenting, regenerating arbitration weights and performing mapping transformation according to actual optimization needs, so as to avoid outputting low-quality content.

[0071] Step 207: When the pre-score value exceeds the preset score threshold, the image to be displayed is shown on the target display interface.

[0072] In this embodiment, the process involves acquiring an image to be processed and performing image preprocessing; performing semantic segmentation on the preprocessed image to obtain semantic segmentation regions contained in the image; generating a heatmap from the preprocessed image to obtain an image heatmap; fusing the semantic segmentation regions and the image heatmap to generate an arbitration weight map; performing a transformation mapping on the arbitration weight map to obtain a transformed image to be displayed; pre-scoring the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value; and displaying the image to be displayed on the target display interface when the pre-scoring value exceeds a preset scoring threshold. This method is used in scenarios where, after capturing or synthesizing a target video, but before displaying it on a real-time screen, image content optimization is performed on each frame of the target video before display. By introducing an arbitration weight map and image quality scoring, it ensures that the displayed image is output with high image quality. Specifically, the arbitration weight map sets a preset inhibitory adjustment mechanism associated with all high-weight regions and a preset enhancement adjustment mechanism associated with all low-weight regions. This ensures that when the arbitration weight map is used to convert the image to be displayed, it avoids over-enhancing the semantic segmentation regions that are more visually noticeable, while ensuring that the image enhancement of semantic segmentation regions that are less visually noticeable is not insufficient, thereby guaranteeing the generation of a high-quality display image.

[0073] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0074] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0075] This application involves acquiring an image to be processed, performing image preprocessing, semantic segmentation of the preprocessed image to obtain semantic segmentation regions contained in the image, generating a heatmap of the preprocessed image, fusing the semantic segmentation regions and the image heatmap to generate an arbitration weight map, performing a transformation mapping on the arbitration weight map to obtain a transformed image to be displayed, pre-scoring the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value, and displaying the image to be displayed on the target display interface when the pre-scoring value exceeds a preset scoring threshold. This method is applicable in scenarios where, after capturing or synthesizing a target video, before displaying the target video in real time on the display interface, image content optimization is performed on each frame of the target video before display. By introducing an arbitration weight map and image quality scoring, it ensures that the displayed image is output with high image quality.

[0076] Further reference Figure 5 As a response to the above Figure 2 The implementation of the method shown in this application provides an embodiment of an image content display optimization device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0077] like Figure 5 As shown, the image content display optimization device 500 described in this embodiment includes: an image acquisition and preprocessing module 501, an image semantic segmentation module 502, an image heatmap generation module 503, an arbitration weight map generation module 504, a displayable image conversion module 505, a displayable image quality scoring module 506, and a displayable image display module 507. Wherein: Image acquisition and preprocessing module 501 is used to acquire the image to be processed and perform image preprocessing. Image semantic segmentation module 502 is used to perform semantic segmentation on the preprocessed image to obtain the semantic segmentation region contained in the image; The image heatmap generation module 503 is used to generate a heatmap from the preprocessed image to obtain an image heatmap. Arbitration weight map generation module 504 is used to fuse the semantic segmentation region and the image heatmap to generate an arbitration weight map; The image conversion module 505 is used to perform conversion mapping on the arbitration weight map to obtain the converted image to be displayed. The image quality scoring module 506 is used to pre-score the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value; The image display module 507 is used to display the image to be displayed on the target display interface when the pre-score value exceeds a preset score threshold.

[0078] This application involves acquiring an image to be processed, performing image preprocessing, semantic segmentation of the preprocessed image to obtain semantic segmentation regions contained in the image, generating a heatmap of the preprocessed image, fusing the semantic segmentation regions and the image heatmap to generate an arbitration weight map, performing a transformation mapping on the arbitration weight map to obtain a transformed image to be displayed, pre-scoring the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value, and displaying the image to be displayed on the target display interface when the pre-scoring value exceeds a preset scoring threshold. This method is applicable in scenarios where, after capturing or synthesizing a target video, before displaying the target video in real time on the display interface, image content optimization is performed on each frame of the target video before display. By introducing an arbitration weight map and image quality scoring, it ensures that the displayed image is output with high image quality.

[0079] In this embodiment, the image acquisition and preprocessing module 501 includes a real-time image extraction unit, a to-be-processed image setting unit, and a color space conversion unit. Wherein: The real-time image extraction unit is used to extract images from video frame sequences or continuous image streams in real time. The image to be processed setting unit is used to set the extracted real-time image as the image to be processed; The color space conversion unit is used to convert the color description of the image to be processed from the initial color space to the target color space according to a preset color space conversion algorithm.

[0080] In this embodiment, the image semantic segmentation module 502 includes a static semantic segmentation unit, a dynamic semantic segmentation unit, and a semantic segmentation region division unit. Wherein: The static semantic segmentation unit is used to perform static semantic segmentation on the preprocessed image using a preset lightweight convolutional neural network to obtain the image regions occupied by different objects in the current real-time image. The dynamic semantic segmentation unit is used to perform dynamic semantic segmentation on the preprocessed image using a preset lightweight convolutional neural network to obtain the image region that has changed dynamically compared to the previous frame image in the current real-time image. The semantic segmentation region division unit is used to divide the image into corresponding semantic segmentation regions based on the image regions occupied by different objects in the current real-time image and the image regions that have dynamically changed compared to the previous frame.

[0081] In this embodiment, the image content display optimization device 500 further includes an inertial attenuation factor allocation module, which is used to allocate a corresponding inertial attenuation factor to each semantic segmentation region according to the semantic segmentation regions contained in the image.

[0082] In this embodiment, the image heatmap generation module 503 includes a gridding processing unit, a visual saliency calculation unit, a thermal characterization value generation unit, and an image heatmap generation unit. Wherein: A gridding processing unit is used to perform gridding processing on the preprocessed image to obtain a series of image grids; The visual saliency calculation unit is used to calculate the attention value for each image grid using the visual saliency calculation method, and to count the attention values ​​of all image grids to generate the attention value distribution. A thermal characterization value generation unit is used to generate thermal characterization values ​​corresponding to all image grids based on the attention value distribution. The image heatmap generation unit is used to generate the image heatmap based on the thermal characterization values ​​corresponding to each of the image grids.

[0083] In this embodiment, the arbitration weight map generation module 504 includes an image grid statistics unit and an arbitration weight map generation unit. Wherein: Image grid statistics unit, used to count the image grid contained in each semantic segmentation region; The arbitration weight map generation unit is used to generate the arbitration weight map by fusing the semantic segmentation region and the image heatmap according to the image grid contained in each semantic segmentation region and the heatmap value corresponding to all image grids, using a preset fusion mode.

[0084] In this embodiment, the image content display optimization device 500 further includes a region weight map acquisition module, a high and low weight region division module, and an adjustment mechanism setting module. Wherein: The region weight map acquisition module is used to set the region weights of all semantic segmentation regions using the preset fusion mode to obtain a region weight map. The high and low weight region division module is used to divide the region weight map into high weight regions and low weight regions according to a preset region weight threshold. The adjustment mechanism setting module is used to set a preset inhibitory adjustment mechanism for all high-weight regions and a preset enhancement adjustment mechanism for all low-weight regions to obtain the arbitration weight map.

[0085] In this embodiment, the region weight map acquisition module includes a first acquisition unit and a second acquisition unit. Wherein: The first obtaining unit is used to calculate the average thermal characterization value of the image grids in each semantic segmentation region based on the image grids contained in each semantic segmentation region and the thermal characterization values ​​corresponding to all image grids, using an averaging algorithm, to obtain the average thermal characterization value corresponding to each semantic segmentation region; and to set the average thermal characterization value corresponding to each semantic segmentation region as the region weight of the corresponding semantic segmentation region to obtain the region weight map. The second obtaining unit is used to calculate the mode thermal representation value of the image grids in each semantic segmentation region based on the image grids contained in each semantic segmentation region and the thermal representation values ​​corresponding to all image grids, using the mode value algorithm to obtain the mode thermal representation value corresponding to each semantic segmentation region; and to set the mode thermal representation value corresponding to each semantic segmentation region as the region weight of the corresponding semantic segmentation region to obtain the region weight map.

[0086] In this embodiment, the image conversion module 505 includes a gamma mapping curve construction unit, a continuous optimization unit, and an image conversion mapping unit. Wherein: The gamma mapping curve construction unit is used to take the region weight of each semantic segmentation region as the input parameter of the baseline gamma curve, and take the adjustment mechanism in the arbitration weight map as the adaptive interpolation compensation mechanism to construct a gamma mapping curve for the image to be displayed to be transformed and mapped by the arbitration weight map. The continuous optimization unit is used to continuously optimize the gamma mapping curve using a continuous optimization method and the adaptive interpolation compensation mechanism until a preset optimization termination condition is met, at which point the continuous optimization stops and the final gamma curve is obtained. The adjustment mechanism in the arbitration weight graph includes the preset inhibitory adjustment mechanism and the preset enhancement adjustment mechanism. The image conversion mapping unit is used to perform image conversion mapping on the arbitration weight map according to the final gamma curve to obtain the converted image to be displayed.

[0087] In this embodiment, the image content display optimization device 500 further includes a feedforward optimization module. The feedforward optimization module is used to trigger a preset parameter rollback or fine-tuning mechanism to perform feedforward optimization when the pre-scoring value does not exceed the preset scoring threshold, until the pre-scoring value exceeds the preset scoring threshold and the feedforward optimization is completed.

[0088] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware with computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0089] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0090] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 6 , Figure 6 This is a basic structural block diagram of the computer device in this embodiment.

[0091] The computer device 6 includes a memory 6a, a processor 6b, and a network interface 6c, which are interconnected via a system bus. It should be noted that... Figure 6 Only a computer device 6 with component memory 6a, processor 6b, and network interface 6c is shown. However, it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Those skilled in the art will understand that the computer device described herein is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0092] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0093] The memory 6a includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 6a may be an internal storage unit of the computer device 6, such as the hard disk or memory of the computer device 6. In other embodiments, the memory 6a may also be an external storage device of the computer device 6, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the computer device 6. Of course, the memory 6a may also include both the internal storage unit and its external storage device of the computer device 6. In this embodiment, the memory 6a is typically used to store the operating system and various application software installed on the computer device 6, such as computer-readable instructions for an image content display optimization method. In addition, the memory 6a can also be used to temporarily store various types of data that have been output or will be output.

[0094] In some embodiments, the processor 6b may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 6b is typically used to control the overall operation of the computer device 6. In this embodiment, the processor 6b is used to execute computer-readable instructions stored in the memory 6a or to process data, for example, to execute computer-readable instructions for the image content display optimization method described above.

[0095] The network interface 6c may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 6 and other electronic devices.

[0096] The computer device proposed in this embodiment belongs to the field of image processing technology and is applied to scenarios where real-time video stream image content is optimized for display using a visual processing model. This application acquires an image to be processed and performs image preprocessing; performs semantic segmentation on the preprocessed image to obtain semantic segmentation regions contained in the image; generates a heatmap of the preprocessed image to obtain an image heatmap; fuses the semantic segmentation regions and the image heatmap to generate an arbitration weight map; performs transformation mapping on the arbitration weight map to obtain a transformed image to be displayed; pre-scores the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value; and displays the image to be displayed on the target display interface when the pre-scoring value exceeds a preset scoring threshold. This method, used in scenarios where image content optimization is performed on each frame of the real-time image in the target video before it is displayed on the display interface after the target video has been captured or synthesized, ensures high-quality output of the displayed image by introducing an arbitration weight map and image quality scoring.

[0097] This application also provides another embodiment, namely, a computer-readable storage medium storing computer-readable instructions that can be executed by a processor to cause the processor to perform the steps of the image content display optimization method described above.

[0098] The computer-readable storage medium proposed in this embodiment belongs to the field of image processing technology and is applied to scenarios where real-time video stream image content is optimized for display using a visual processing model. This application acquires an image to be processed and performs image preprocessing; performs semantic segmentation on the preprocessed image to obtain semantic segmentation regions contained in the image; generates a heatmap of the preprocessed image to obtain an image heatmap; fuses the semantic segmentation regions and the image heatmap to generate an arbitration weight map; performs transformation mapping on the arbitration weight map to obtain a transformed image to be displayed; pre-scores the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value; and displays the image to be displayed on the target display interface when the pre-scoring value exceeds a preset scoring threshold. This method is used in scenarios where image content optimization is performed on each frame of the real-time image in the target video before it is displayed on the display interface after the target video has been captured or synthesized. By introducing an arbitration weight map and image quality scoring, it ensures that the displayed image is output with high image quality.

[0099] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0100] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to make the disclosure of this application more thorough and comprehensive. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application. Software tools or components not belonging to this company that appear in the embodiments of this application are merely illustrative examples and do not represent actual use.

Claims

1. A method for optimizing image content display, characterized in that, Includes the following steps: Acquire the image to be processed and perform image preprocessing; Semantic segmentation is performed on the preprocessed image to obtain the semantic segmentation regions contained in the image; A heatmap is generated from the preprocessed image to obtain an image heatmap; By fusing the semantic segmentation region and the image heatmap, an arbitration weight map is generated; The arbitration weight graph is transformed and mapped to obtain the transformed image to be displayed; The image to be displayed is pre-scored according to a preset image quality scoring component to obtain a pre-scoring value; When the pre-score exceeds the preset score threshold, the image to be displayed is shown on the target display interface.

2. The image content display optimization method according to claim 1, characterized in that, The steps of acquiring the image to be processed and performing image preprocessing specifically include: Real-time image extraction from video frame sequences or continuous image streams; The extracted real-time image is used as the image to be processed; According to a preset color space conversion algorithm, the color description of the image to be processed is converted from the initial color space to the target color space.

3. The image content display optimization method according to claim 1, characterized in that, The semantic segmentation regions contained in the image include static segmentation regions corresponding to different objects in the same frame and dynamic change regions between consecutive frames. The step of performing semantic segmentation on the preprocessed image to obtain the semantic segmentation regions contained in the image specifically includes: A pre-defined lightweight convolutional neural network is used to perform static semantic segmentation on the preprocessed image, obtaining the image regions occupied by different objects in the current real-time image, and A pre-defined lightweight convolutional neural network is used to perform dynamic semantic segmentation on the preprocessed image to obtain the image region that has dynamically changed compared to the previous frame image in the current real-time image. Based on the image regions occupied by different objects in the current real-time image and the image regions that have dynamically changed compared to the previous frame, corresponding semantic segmentation regions are divided in the image.

4. The image content display optimization method according to claim 1, characterized in that, The step of generating a heatmap from the preprocessed image to obtain an image heatmap specifically includes: The preprocessed image is then subjected to a gridding process to obtain a series of image grids; The attention value is calculated for each image grid using a visual saliency calculation method. The attention values ​​for all image grids are then statistically analyzed to generate the attention value distribution. Based on the attention value distribution, generate thermal characterization values ​​corresponding to each of the image grids; The image heatmap is generated based on the thermal characterization values ​​corresponding to each of the image grids.

5. The image content display optimization method according to claim 4, characterized in that, The step of fusing the semantic segmentation region and the image heatmap to generate the arbitration weight map specifically includes: Analyze the image grid contained in each semantic segmentation region; Based on the image grids contained in each semantic segmentation region and the heatmap values ​​corresponding to all image grids, a preset fusion mode is used to fuse the semantic segmentation regions and the image heatmaps to generate the arbitration weight map.

6. The image content display optimization method according to claim 5, characterized in that, The step of generating the arbitration weight map by fusing the semantic segmentation region and the image heatmap using a preset fusion mode includes: Using the preset fusion mode, the region weights of all semantic segmentation regions are set to obtain a region weight map; Based on a preset regional weight threshold, the region weight map is divided into high-weight regions and low-weight regions. The arbitration weight map is obtained by setting a pre-defined inhibitory adjustment mechanism for all high-weight regions and a pre-defined enhancement adjustment mechanism for all low-weight regions.

7. The image content display optimization method according to claim 6, characterized in that, The step of transforming and mapping the arbitration weight map to obtain the transformed image to be displayed specifically includes: The region weight of each semantic segmentation region is used as the input parameter of the baseline gamma curve, and the adjustment mechanism in the arbitration weight map is used as the adaptive interpolation compensation mechanism to construct a gamma mapping curve for the image to be displayed to be transformed and mapped by the arbitration weight map. The gamma mapping curve is continuously optimized using a continuous optimization method and the adaptive interpolation compensation mechanism until a preset optimization termination condition is met, at which point the continuous optimization stops and the final gamma curve is obtained. The adjustment mechanism in the arbitration weight graph includes the preset inhibitory adjustment mechanism and the preset enhancement adjustment mechanism. Based on the final gamma curve, the arbitration weight map is transformed and mapped to obtain the transformed image to be displayed.

8. An image content display optimization device, characterized in that, include: The image acquisition and preprocessing module is used to acquire the image to be processed and perform image preprocessing. The image semantic segmentation module is used to perform semantic segmentation on the preprocessed image to obtain the semantic segmentation regions contained in the image; An image heatmap generation module is used to generate a heatmap from the preprocessed image to obtain an image heatmap. The arbitration weight map generation module is used to fuse the semantic segmentation region and the image heatmap to generate an arbitration weight map; The image conversion module is used to convert and map the arbitration weight map to obtain the converted image to be displayed. The image quality scoring module is used to pre-score the image to be displayed according to a preset image quality scoring component to obtain a pre-scoring value; The image to be displayed module is used to display the image to be displayed on the target display interface when the pre-score value exceeds a preset score threshold.

9. A computer device, characterized in that, The method includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the image content display optimization method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the image content display optimization method as described in any one of claims 1 to 7.

Images