MLED Evaluation Method and System Integrating Standard Image Database and Dynamic Image Calibration

Abstract

This invention relates to the field of LED display evaluation technology, solving the problems of low accuracy and inability to perform multi-dimensional evaluation of MLED display quality in existing technologies. It provides an MLED evaluation method and system that integrates a standard image database and dynamic image calibration. The method includes: acquiring a standard image database corresponding to multiple evaluation items of the MLED display terminal to be evaluated; controlling the MLED display terminal to display target standard images from the standard image database, and acquiring real-time image data corresponding to the target standard images displayed by the MLED display terminal; evaluating the MLED display terminal according to a preset evaluation strategy corresponding to each evaluation item, combined with the real-time image data and the target standard images, to obtain a target evaluation result corresponding to the display quality of the MLED display terminal. This invention improves the accuracy of MLED evaluation and solves the problem of single evaluation dimensions in existing technologies.

Description

Technical Field

[0001] This invention relates to the field of LED display evaluation, and in particular to an MLED evaluation method and system that integrates a standard image database and dynamic image calibration. Background Technology

[0002] MLED is a collective term for Mini LED and Micro LED, both of which belong to the next generation of self-emissive display technologies. Mini LED achieves higher dynamic contrast and brightness control precision by miniaturizing the size of traditional LED chips and increasing the number of backlight zones. Micro LED uses micron-level LED chips as pixel units to achieve true pixel-level self-emissiveness, ultra-high brightness, long lifespan, and low energy consumption. Due to its high contrast, wide color gamut, high response speed, ultra-low latency, and excellent environmental adaptability, MLED technology is widely used in high-end TVs, automotive displays, wearable devices, AR / VR headsets, and indoor ultra-high-definition displays.

[0003] Due to the complex structure, high pixel density, and sophisticated manufacturing process of MLED display terminals, products are prone to display defects such as uneven brightness, color shift, decreased sharpness, and viewing angle dependence during actual production and application. Therefore, to ensure that MLED display devices meet the expected display quality standards before leaving the factory, it is necessary to establish a systematic, scientific, and multi-dimensional display quality evaluation mechanism. This mechanism not only helps identify quality problems in the manufacturing process but also supports product grading, user experience optimization, and adaptability assessment for specific application scenarios.

[0004] Chinese patent CN115616787A discloses a screen display calibration assistance method and a computer-readable storage medium. The method includes: acquiring a test chart corresponding to calibration items of a head-mounted device to be calibrated; sending the test chart to the head-mounted device for display on its screen; capturing images of the test chart displayed on the head-mounted device screen using a lens simulating the optical structure of the human eye to obtain first captured image data; analyzing the first captured image data according to an analysis strategy corresponding to the calibration items to obtain calibration assistance data; and sending the calibration assistance data to the head-mounted device for screen display calibration based on the calibration assistance data. This patent primarily focuses on the objective quality assessment of terminal display effects. The screen display calibration assistance method relies heavily on manually selected test charts for each calibration item and manually set shooting and analysis processes, providing one-way calibration data transmission and feedback to the head-mounted device. This method has two main technical problems in the calibration process: First, the calibration process and the evaluation process do not form a closed loop linkage, and there is a lack of real-time quantitative verification and strategy adjustment of the calibration effect; Second, it has weak adaptability under various environmental variables (such as user viewing angle shift, wearing differences, and changes in ambient light), making it difficult to achieve high-precision, adaptive, and personalized calibration, thus affecting the consistency and reliability of the final display quality.

[0005] Therefore, how to evaluate the quality of MLED displays from multiple dimensions and improve the accuracy of the evaluation is an urgent technical problem to be solved. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide an MLED evaluation method and system that integrates a standard image database and dynamic image calibration, in order to solve the problem that the existing technology cannot perform multi-dimensional evaluation of MLED display quality and has low evaluation accuracy.

[0007] In a first aspect, embodiments of the present invention provide an MLED evaluation method that integrates a standard image database and dynamic image calibration, the method comprising:

[0008] Obtain a standard image database corresponding to multiple evaluation items of the MLED display terminal to be evaluated;

[0009] The MLED display terminal is controlled to display the target standard image in the standard image database, and real-time image data corresponding to the target standard image displayed by the MLED display terminal is acquired.

[0010] Based on the preset evaluation strategy corresponding to each evaluation item, and in combination with the real-time image data and the target standard image, the MLED display terminal is evaluated to obtain the target evaluation result corresponding to the display quality of the MLED display terminal.

[0011] Preferably, the standard image database includes at least one of the following test images: uniformity test image, linearity test image, distortion test image, brightness test image, chromaticity test image, display resolution test image, and crosstalk test image.

[0012] Preferably, the standard image database corresponding to multiple evaluation items of the MLED display terminal to be evaluated includes:

[0013] Obtain the preset initial weights of each test image in each of the evaluation items;

[0014] Based on the terminal type of the MLED display terminal, and combined with the preset mapping relationship between the terminal type and the weight correction factor, a first weight correction factor corresponding to the terminal type is obtained.

[0015] Based on the usage scenario type of the MLED display terminal, and combined with the preset mapping relationship between the usage scenario type and the weight correction factor, a second weight correction factor corresponding to the usage scenario type is obtained;

[0016] The initial weights are corrected based on the first weight correction factor and the second weight correction factor to obtain the target weights of each test image in each of the evaluation sub-items.

[0017] Based on the target weights, the test images corresponding to each evaluation item are determined as the standard image databases for each item.

[0018] Preferably, determining the test images corresponding to each of the evaluation items as the standard image databases based on each of the target weights includes:

[0019] Obtain the ambient illuminance and chromaticity values ​​of the MLED display terminal.

[0020] Based on the ambient illuminance of the test environment, the evaluation adaptability of each test image under the current ambient illuminance conditions is analyzed to obtain the illuminance adaptability factor.

[0021] Based on the chromaticity values ​​of the test environment, the evaluation adaptability of each test image under the current chromaticity value conditions is analyzed to obtain the chromaticity adaptation factor;

[0022] Based on the illuminance adaptation factor and chromaticity adaptation factor corresponding to each test image, the weights of each target are weighted and calculated to obtain the adaptation weights of each environment.

[0023] The environmental adaptation weights of each of the evaluation items are sorted, and the standard image databases are determined based on the sorting results.

[0024] Preferably, the step of analyzing the suitability of each test image under the current ambient light conditions based on the ambient light intensity to obtain the light intensity suitability factor includes:

[0025] The test environment illuminance is mapped according to the preset illuminance level range to obtain the target level range corresponding to the test environment illuminance.

[0026] Classify each test image in the standard image database to obtain the image category of each test image;

[0027] Based on the preset mapping relationship between image category and light sensitivity level, the target light sensitivity level corresponding to each test image is obtained;

[0028] The illuminance of the test environment is analyzed for fluctuation to obtain the illuminance stability, wherein the illuminance stability includes the illuminance change rate and the illuminance fluctuation frequency;

[0029] For the target level range, the target illuminance sensitivity level and the illuminance stability are weighted and fused to obtain the illuminance adaptation factor.

[0030] Preferably, the step of analyzing the adaptability of each test image under the current environmental chromaticity value based on the test environment chromaticity value to obtain the chromaticity adaptation factor includes:

[0031] The deviation between the test environment chromaticity value and the preset standard chromaticity value is calculated to obtain the chromaticity deviation value;

[0032] Based on the chromaticity deviation value, determine the direction and level of chromaticity deviation;

[0033] The chromaticity adaptation factor is obtained by weighting the chromaticity deviation direction, the chromaticity deviation level, and the preset chromaticity sensitivity corresponding to each test image.

[0034] Preferably, the step of evaluating the MLED display terminal based on a preset evaluation strategy corresponding to each of the evaluation items, combined with the real-time image data and the target standard image, to obtain a target evaluation result corresponding to the display quality of the MLED display terminal includes:

[0035] Based on the terminal type and the usage scenario type, determine the evaluation priority corresponding to each evaluation item;

[0036] Based on the evaluation priorities described above, determine the target priorities that meet the preset priority requirements, and obtain the target evaluation items corresponding to the target priorities;

[0037] According to the preset evaluation strategy, obtain the target evaluation strategy corresponding to the target evaluation item;

[0038] Based on the target evaluation strategy, and combining the real-time image data and the target standard image, the MLED display terminal is evaluated to obtain a first evaluation sub-result;

[0039] Based on the first evaluation sub-result and the preset evaluation early termination condition, determine whether the first evaluation sub-result meets the evaluation early termination condition;

[0040] If the first evaluation sub-result meets the early termination condition of the evaluation, then the first evaluation sub-result will be directly used as the target evaluation result.

[0041] If the first evaluation sub-result does not meet the early termination condition of the evaluation, then according to the evaluation priority, the evaluation strategy corresponding to the non-target evaluation item in each evaluation item is obtained in sequence.

[0042] Based on the evaluation strategy corresponding to the non-target evaluation items, and combined with the real-time image data and the target standard image, the MLED display terminal is evaluated sequentially to obtain each second evaluation sub-result;

[0043] Based on the first evaluation sub-result and in combination with each of the second evaluation sub-results, the target evaluation result is determined.

[0044] Preferably, determining whether the first evaluation sub-result meets the early evaluation termination condition based on the first evaluation sub-result and the preset early evaluation termination condition includes:

[0045] The target evaluation items are classified to determine the target evaluation item type, and the parameter threshold set corresponding to the early termination condition of the evaluation for that target item type is obtained.

[0046] The evaluation parameters are extracted from the first evaluation sub-result to obtain a set of key evaluation parameters;

[0047] Based on the evaluation logic type corresponding to each parameter threshold in the parameter threshold set, type mapping and normalization processing are performed on each evaluation parameter in the key evaluation parameter set to obtain the standard parameter set corresponding to each parameter threshold.

[0048] Based on the correspondence between the set of standard parameters and the set of parameter thresholds, each standard parameter value is compared with its corresponding parameter threshold one by one to obtain a set of comparison results for each standard parameter.

[0049] Based on the set of comparison results, determine whether the first evaluation sub-result meets the early termination condition of the evaluation.

[0050] Preferably, determining the target evaluation result based on the first evaluation sub-result and in combination with the second evaluation sub-result includes:

[0051] Obtain the first evaluation weight corresponding to the target evaluation item and the second evaluation weight corresponding to each of the non-target evaluation items;

[0052] Based on each of the second evaluation weights, the second evaluation results are weighted and calculated to obtain the initial evaluation results;

[0053] The initial evaluation result and the first evaluation result are weighted and calculated based on the first evaluation weight and each of the second evaluation weights to obtain the target evaluation result.

[0054] In a second aspect, embodiments of the present invention provide an MLED evaluation system that integrates a standard image database and dynamic image calibration. The system includes at least one processor, at least one memory, and computer program instructions stored in the memory. When the computer program instructions are executed by the processor, they implement the method described in the first aspect.

[0055] In summary, the beneficial effects of the present invention are as follows:

[0056] This invention provides an MLED evaluation method and system that integrates a standard image database and dynamic image calibration. The method includes: acquiring a standard image database corresponding to multiple evaluation items of an MLED display terminal to be evaluated; controlling the MLED display terminal to display a target standard image from the standard image database, and acquiring real-time image data corresponding to the target standard image displayed by the MLED display terminal; evaluating the MLED display terminal according to a preset evaluation strategy corresponding to each evaluation item, combined with the real-time image data and the target standard image, to obtain a target evaluation result corresponding to the display quality of the MLED display terminal. This invention constructs a comprehensive evaluation mechanism for dynamic perception and multi-dimensional adaptation of MLED display terminals, effectively solving the technical problems of traditional evaluation methods, such as the lack of scene-specificity in standard image selection, the sensitivity of evaluation results to ambient light, and the inability to provide real-time feedback on display status during the evaluation process. First, this invention introduces a standard image database containing various typical test images and adaptively corrects the image weights based on terminal type, usage scenario, ambient light level, and color temperature conditions to ensure highly adaptable evaluation image selection. Second, by controlling the MLED terminal to dynamically display the target standard image and collect real-time image data, the evaluation process reflects the actual display effect of the terminal in real time, enhancing the synchronization between calibration accuracy and evaluation. Third, by setting multi-dimensional preset evaluation strategies, evaluation priorities, and early termination mechanisms, rapid judgment of key quality indicators and supplementary judgment of secondary indicators are achieved, balancing efficiency and comprehensiveness. Finally, by calculating the results of each sub-item through hierarchical weighted calculation and fusing the first and second evaluation sub-results, a precise and reliable comprehensive evaluation result is output. Therefore, this invention improves the accuracy, consistency, and environmental robustness of evaluation while achieving closed-loop control from standard images to dynamic evaluation results, comprehensively solving the problems of difficult, low-precision, and slow feedback in evaluating the display quality of MLED display terminals in complex usage environments. Attached Figure Description

[0057] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, and these are all within the protection scope of the present invention.

[0058] Figure 1 This is a schematic diagram of the overall process of the MLED evaluation method that integrates a standard image database and dynamic image calibration in Embodiment 1 of the present invention;

[0059] Figure 2 This is a schematic diagram of the process for evaluating the MLED display terminal in Embodiment 1 of the present invention;

[0060] Figure 3 This is a schematic diagram of the structure of the MLED evaluation system that integrates a standard image database and dynamic image calibration in Embodiment 2 of the present invention. Detailed Implementation

[0061] The features and exemplary embodiments of various aspects of the present invention will now be described in detail. To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present invention and are not configured to limit the present invention. For those skilled in the art, the present invention can be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present invention by illustrating examples of the invention.

[0062] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0063] It should be noted that all actions involving the acquisition of signals, information, or data in this invention are carried out in compliance with the relevant data protection laws and regulations of the locality and with authorization from the owner of the relevant device.

[0064] Example 1

[0065] Please see Figure 1 This invention provides an MLED evaluation method that integrates a standard image database and dynamic image calibration. The method includes:

[0066] Obtain a standard image database corresponding to multiple evaluation items of the MLED display terminal to be evaluated;

[0067] Specifically, the standard image database refers to a set of professionally designed images used to test display performance characteristics, including but not limited to uniform grayscale images, color bar images, resolution grid images, and geometric image images. Evaluation items refer to multiple specific technical dimensions used to measure the quality of MLED display terminals, such as brightness consistency, contrast ratio, grayscale performance, color accuracy, and image distortion. The purpose of this step is to match and select representative standard images as evaluation benchmarks based on the chosen evaluation items, ensuring the relevance and scientific rigor of subsequent evaluations. The process of implementing this step includes: first, identifying multiple key indicators required for evaluating the current MLED display terminal; then, filtering images from a preset image library according to the image type required for each evaluation item, and constructing a target standard image database containing multiple test images; in more complex implementations, the selection weight of standard images can be dynamically adjusted based on parameters such as terminal type, usage scenario, lighting, and color temperature environment to ensure high adaptability of image content. The technical feature of this step is that it achieves structured image management through a project-image one-to-one mapping mechanism, which has good scalability and flexibility. It ensures that the standard images used have high representativeness and testing efficiency, laying a reliable foundation for subsequent dynamic image comparison and display quality evaluation, and improving the accuracy and objectivity of the evaluation from the source.

[0068] The MLED display terminal is controlled to display the target standard image in the standard image database, and real-time image data corresponding to the target standard image displayed by the MLED display terminal is acquired.

[0069] Specifically, real-time image data refers to the image record of the terminal display effect, synchronously acquired from the observation angle while the MLED display terminal outputs an image through acquisition devices such as cameras and image sensors. The purpose of this step is to accurately project the standard image onto the terminal screen and objectively acquire the actual display effect image, thus providing a reliable input for subsequent image comparison and quality evaluation. The MLED terminal is controlled to display the target standard images sequentially according to a preset time sequence. The duration of each image meets the sensor's exposure and sampling requirements. Simultaneously, the acquisition device captures the actual image displayed by the terminal from a fixed observation position. Further processing such as illumination compensation, background removal, and reflection reduction can be considered during the acquisition process to ensure data cleanliness and consistency. This achieves a one-to-one correspondence between the standard signal image and the actual display image, exhibiting high temporal synchronization and spatial consistency. This enables the system to accurately obtain the terminal's true display effect, significantly improving the reliability of subsequent comparison and evaluation, and avoiding evaluation errors caused by subjective observation or screen differences.

[0070] Based on the preset evaluation strategy corresponding to each evaluation item, and in combination with the real-time image data and the target standard image, the MLED display terminal is evaluated to obtain the target evaluation result corresponding to the display quality of the MLED display terminal.

[0071] Specifically, the pre-set evaluation strategy refers to the specific analysis rules, calculation methods, and evaluation criteria formulated for each evaluation item. Through this strategy, standard images and acquired images are objectively compared, and the actual display performance of the terminal is quantified using technical means to output the target evaluation result. The implementation process includes: image registration, alignment, and normalization of the acquired real-time images and standard images; calculation of evaluation index values ​​according to each evaluation strategy; and weighted fusion of multiple indicators using weighting coefficients to ultimately form a comprehensive target evaluation result. In advanced implementations, image recognition algorithms can be introduced to perform local quality analysis on key areas, improving evaluation accuracy. By adopting a multi-dimensional index fusion mechanism based on image comparison, an automated, standardized, and repeatable evaluation process is achieved, significantly reducing human subjective error and enabling high-precision, high-efficiency, and comprehensive evaluation of MLED display performance, adapting to the needs of large-scale terminal quality inspection and performance optimization.

[0072] Preferably, the standard image database includes at least one of the following test images: uniformity test image, linearity test image, distortion test image, brightness test image, chromaticity test image, display resolution test image, and crosstalk test image.

[0073] Specifically, an image test set covering multi-dimensional display performance indicators is constructed to support a comprehensive quality evaluation of MLED display terminals. The various image types are designed for specific evaluation items: uniformity test images are typically large-area grayscale images used to detect whether the screen brightness or color distribution is uniform; linearity test images contain grayscale or color gradation transition blocks, reflecting the linear response capability of brightness or color output; distortion test images are mostly regular graphics (such as grids or circles) used to reveal geometric distortion phenomena; brightness test images are blocks of different brightness levels used to analyze the screen brightness response range; chromaticity test images are combinations of standard color blocks used for color accuracy and color reproduction testing; display resolution test images often use multi-frequency stripes or dot matrix patterns to detect the terminal's sharpness and resolution limits; crosstalk test images are used to evaluate the interference effect of displayed content in adjacent areas, particularly suitable for crosstalk problems in multi-channel or microlens array structures. The purpose of this step is to construct a multi-type test image set, thereby ensuring the standard image database has a unified breadth and depth of evaluation. This provides accurate test samples for each display performance indicator, avoiding the risk of a particular type of image failing or causing biased results in multiple evaluations. This meticulously categorized and comprehensive image library design ensures that the subsequent evaluation system can select the optimal image set when facing different terminal structures and application scenarios, enhancing the system's versatility and accuracy. In implementation, high-quality image samples conforming to industrial testing standards are first constructed using image design tools. Each image type must have a unified format, resolution, color encoding, and other parameters. Next, based on the evaluation purpose of each test image, it is labeled and categorized, and entered into the standard image database system for subsequent on-demand retrieval. Furthermore, image version management and parameter labeling can be implemented, such as variant images with different brightness levels, different grayscale densities, and different color gamuts, to enhance the database's flexible adaptability. The database also supports extended functions such as multilingual tags, evaluation item mapping, and terminal type adaptation strategies, improving its practicality.

[0074] Preferably, the standard image database corresponding to multiple evaluation items of the MLED display terminal to be evaluated includes:

[0075] Obtain the preset initial weights of each test image in each of the evaluation items;

[0076] Specifically, based on the characteristics and usage scenarios of the MLED display terminals to be evaluated, the priority of each test image in the standard image database is dynamically adjusted to ensure that the evaluation strategy has differentiated adaptability and practical applicability. First, "obtaining the preset initial weights of each test image in each evaluation item" means that for multiple evaluation items such as brightness, chromaticity, and resolution, the system assigns a default weight value to each type of test image. For example, a brightness test image has a weight of 0.8 in brightness evaluation and only 0.2 in resolution evaluation. These initial weights reflect the original importance of each test image in each evaluation dimension and are the basic configuration for building the standard image database. The purpose of this step is to establish a basic weight system, providing a starting point for subsequent personalized weight adjustments based on specific terminal conditions, thereby realizing adjustments to the testing focus due to terminal differences.

[0077] Based on the terminal type of the MLED display terminal, and combined with the preset mapping relationship between the terminal type and the weight correction factor, a first weight correction factor corresponding to the terminal type is obtained.

[0078] Specifically, terminal type refers to the specific application category of the MLED display terminal, such as outdoor advertising screens, vehicle HUDs, conference screens, home theater TVs, and stage background screens. Different terminal types exhibit significant differences in structural characteristics, environmental exposure, and viewing distance. The weight correction factor is a correction coefficient set to adjust the evaluation weight of the test image under a specific terminal type. For example, in vehicle HUDs, brightness uniformity and anti-glare capability are crucial, and the corresponding test image weight correction factor might be 0.3, while in home theater scenarios, color saturation is more important, and the corresponding image correction factor might be 0.4. Introducing the terminal type as an influencing factor, through a differentiated weighting mechanism, ensures that the standard image database selected and used in subsequent evaluation processes better matches the actual application needs of the terminal. Since different types of MLED terminals differ significantly in their intended use, viewing environment, installation method, and target user experience, using a single universal image weighting system could easily lead to generalized or even distorted evaluation results. Therefore, the logical derivation of this step is: by pre-establishing a mapping model between terminal type and correction factor, the image weights acquire terminal structure awareness capabilities to support the construction of personalized testing standards. First, the system incorporates or extracts a terminal type-evaluation item weight correction table from an empirical model. This table can be generated through training based on industry standards, expert experience, or a large amount of real-world data. It includes the weight offset for each type of terminal in each evaluation dimension. For example, the table specifies a correction factor of +0.1 for outdoor large screens on chromaticity test images and +0.3 on resolution test images. Second, the system reads the type information of the terminal under test. This information can be input by the user, automatically identified (e.g., by reading the terminal model via an interface), or preset by the testing platform. Finally, the system retrieves the correction factor for the corresponding terminal type from this mapping table and outputs the first set of weight correction factors for that terminal type, which is used for subsequent calculations of the target weight correction.

[0079] Based on the usage scenario type of the MLED display terminal, and combined with the preset mapping relationship between the usage scenario type and the weight correction factor, a second weight correction factor corresponding to the usage scenario type is obtained;

[0080] Specifically, the usage scenario type refers to the application environment or functional background in which the MLED display terminal is actually deployed and operated, such as stadiums, museum displays, outdoor advertising, vehicle displays, home theaters, and conference room displays. Different scenarios have different focuses on display performance. For example, outdoor advertising emphasizes brightness penetration, while museum scenarios prioritize color reproduction and low blue light comfort. The "second weight correction factor" is an image evaluation weight adjustment parameter designed to adjust the weight of each image in the image evaluation, based on the impact of the usage scenario on the importance of various test images. This parameter dynamically corrects the weight of each image in the image evaluation, aiming to achieve application scenario awareness of the evaluation weight, thereby further improving the practicality and relevance of the test results. In practical applications, the same model of MLED display terminal may be deployed in drastically different environments. For example, the same panel may be used for public transportation information display and art exhibition display, and the evaluation focus will obviously be different. If the evaluation system ignores the variable of usage scenario, the results will deviate from the actual user experience, reducing the reference value of the evaluation. Therefore, the logical derivation of this step is as follows: By constructing a mapping model of scene type and correction factor, the evaluation system can flexibly adjust the evaluation weight of each test image according to the usage scenario, making the evaluation system closer to the essence of the application. A mapping table or rule model between usage scenario type and evaluation item weight correction factor is constructed and stored. This model can be formed based on a large amount of real-world user feedback, industry standard data, and scenario simulation experiments. For example, in a sports stadium scenario, the weight correction for dynamic response images is 0.25, and in a home theater scenario, the correction for chromaticity test images is 0.4. The usage scenario type of the current MLED display terminal is determined through user interface input, deployment of scene recognition, and calling configuration files. Based on the current usage scenario type, the set of weight correction factors corresponding to that scenario is retrieved from the preset mapping model and used as the second weight correction factor. This factor set is provided to subsequent image weight correction steps to support the calculation of the final weight value. Introducing a dynamic adjustment mechanism of the scene dimension during the image weight correction process ensures that the participation of test images is not only related to the terminal type but also strongly correlated with the actual application environment in which the terminal is located. Firstly, it enhances the context-awareness of the evaluation system, enabling evaluation results to reflect actual user experience. Secondly, it strengthens the generalizability of the evaluation model, making the system widely applicable to diverse and personalized MLED deployment needs. Thirdly, it optimizes image resource allocation strategies to avoid invalid calculations caused by the participation of low-relevance images in the evaluation, thereby improving evaluation efficiency and accuracy and providing more scenario-adaptable technical support for terminal quality control, customization optimization, and delivery acceptance.

[0081] The initial weights are corrected based on the first weight correction factor and the second weight correction factor to obtain the target weights of each test image in each of the evaluation sub-items.

[0082] Specifically, firstly, the initial weights refer to the pre-defined proportions of importance of each test image in different evaluation sub-items (such as brightness uniformity, chromaticity consistency, contrast boundary, etc.) under ideal or general scenarios. The first and second weight correction factors are extracted based on the hardware type of the MLED terminal (such as small-pitch LED, transparent LED, automotive LED, etc.) and specific usage scenarios (such as cinema display, in-vehicle navigation, outdoor advertising, etc.). These factors are essentially dynamic adjustment coefficients that quantify the changes in the evaluation value of test images under different conditions. The aim is to achieve fine-tuning of the test image weights in both the "end-field" and "end-to-end" dimensions based on the general initial weights, making the image composition in the final evaluation database more closely match the actual operating state and display requirements of specific MLED terminals, thereby improving the accuracy of the subsequent evaluation model in judging the true display quality. Its core logic is the superposition and fusion of weight correction functions. By introducing coefficient parameters that are bidirectionally related to terminal type and usage scenario, the weight model is given dynamic adaptability. In the implementation process, the initial weights of each test image in each evaluation sub-project are first extracted. Then, the correction factor sets corresponding to the terminal type and usage scenario are called, and item-by-item weighted corrections are performed accordingly. Common forms include linear weighting or non-linear function correction (such as exponential or two-factor interpolation models). After correction, the target weight matrix under the current terminal and scenario is obtained. This target weight set is then used as the basis for constructing a standard image database to guide the participation ratio of different test images in specific evaluation projects.

[0083] Based on the target weights, the test images corresponding to each evaluation item are determined as the standard image databases for each item.

[0084] Specifically, the target weights are importance parameters for each test image in each evaluation sub-item, obtained in the previous step based on the initial weights and combined with the terminal type and usage scenario type of the MLED display terminal. Test images refer to a preset image set used to evaluate display performance, such as uniformity test images, brightness test images, and distortion test images. Evaluation items include multiple dimensions such as brightness consistency, color fidelity, resolution performance, and dynamic response capability. Based on the quantified target weights, the most representative, sensitive, and suitable images for each evaluation item are selected from the complete test image set to form a standard image database. This process ensures that the image basis upon which the evaluation relies is highly consistent with the performance characteristics and usage environment conditions of the current MLED terminal, thereby ensuring the reliability and effectiveness of the features extracted from each evaluation sub-item in subsequent evaluation processes. In the specific implementation process, based on the preset mapping relationship between each evaluation item and the test image, and combined with the corrected target weight set, image filtering and matching operations are performed. The filtering method can employ a weighted sorting method, where all candidate test images are sorted in descending order according to their target weight for a given evaluation item, and the top few images with the highest weight values ​​are selected as the standard images for that evaluation item. Alternatively, a weight threshold filtering method can be used, retaining only test images with weights greater than a set threshold for constructing the standard image database. Furthermore, in specific scenarios, additional dimensions such as environmental adaptability factors and historical image evaluation stability can be incorporated to optimize the image selection process, further improving the representativeness and applicability of the images. The standard image database constructed through this step possesses strong targeting and flexibility. Its technical features bring beneficial effects including: firstly, improving the correlation between the image input data and evaluation indicators relied upon by various evaluation tasks, avoiding evaluation bias due to "images not matching the topic"; secondly, by dynamically filtering and constructing the standard image database, effectively reducing image redundancy, saving system resources, improving evaluation efficiency, and laying a solid data foundation for subsequent automated and intelligent evaluation processes.

[0085] Preferably, determining the test images corresponding to each of the evaluation items as the standard image databases based on each of the target weights includes:

[0086] Obtain the ambient illuminance and chromaticity values ​​of the MLED display terminal.

[0087] Specifically, firstly, the ambient illuminance and chromaticity values ​​of the MLED display terminal are acquired. Illuminance refers to the ambient brightness intensity under test conditions, typically measured in Lux; chromaticity values ​​are tristimulus values ​​(e.g., CIE XYZ) or color space coordinates (e.g., CIE Lab, CIE xyY). The purpose of this step is to collect ambient light parameters under the current test conditions to dynamically adjust the image selection strategy and improve the practical applicability of the evaluation. In practice, an ambient light sensor and colorimeter are typically configured to collect and store the illuminance and color data of the test scene in real time, providing an environmental benchmark for subsequent image selection. The key technical feature of this step is that it uses objective ambient light parameters as a basis to provide raw input for adaptive image evaluation, effectively improving the system's intelligent environmental perception capabilities and ensuring consistency and accuracy in the evaluation process under different lighting and color conditions.

[0088] Based on the ambient illuminance of the test environment, the evaluation adaptability of each test image under the current ambient illuminance conditions is analyzed to obtain the illuminance adaptability factor.

[0089] Specifically, the illuminance adaptation factor is an indicator that quantifies the visual stability and evaluation sensitivity of test images under current lighting conditions. The purpose of this step is to determine which test images can still accurately represent their design evaluation characteristics under the current brightness background. For example, in a high-brightness environment, the boundaries of low-contrast images may be blurred, making them unsuitable for resolution evaluation. The implementation process typically involves mapping the current test environment's illuminance to multiple preset level ranges (such as low light, medium light, and strong light), and combining this with the corresponding illuminance sensitivity level of each image to calculate the illuminance adaptation factor for each image using a weighted model. This step, by introducing modeling of the impact of the lighting environment on image expressiveness, effectively avoids evaluation distortion caused by incompatible lighting environments and enhances the adaptability of the evaluation strategy to actual usage conditions.

[0090] Based on the chromaticity values ​​of the test environment, the evaluation adaptability of each test image under the current chromaticity value conditions is analyzed to obtain the chromaticity adaptation factor;

[0091] Specifically, the chromaticity adaptation factor reflects the degree of interference of environmental color deviation on image evaluation and the image's own color resistance. Its purpose is to identify test images that still possess strong color expression accuracy under current environmental color deviation conditions. The implementation involves: first, calculating the deviation between the current chromaticity value and a preset standard chromaticity (such as D65 standard white light); then, constructing an evaluation adaptation model based on the image's chromaticity sensitivity level and deviation direction (blue-biased, red-biased, green-biased, etc.) to obtain an adaptation score for each image under the current color deviation. This step, by modeling the coupling relationship between environmental color deviation and image sensitivity, further strengthens the environmental matching capability of the test image selection process, ensuring that the finally selected test images maintain a high degree of consistency in both subjective visual observation and objective imaging evaluation.

[0092] Based on the illuminance adaptation factor and chromaticity adaptation factor corresponding to each test image, the weights of each target are weighted and calculated to obtain the adaptation weights of each environment.

[0093] Specifically, after obtaining the illuminance and chromaticity adaptation factors for each test image, the system performs a weighted calculation on each target weight based on these factors to obtain the environmental adaptation weights. The target weights themselves reflect the theoretical importance of the test image in different evaluation sub-items, while the environmental adaptation factors embody the image's expressive power in the current test environment. Therefore, this weighted calculation step aims to integrate the image's evaluation value with its usability in the current test environment to construct a dynamic weight system. In implementation, the system typically uses a weighted average or normalized weighting mechanism to integrate the illuminance and chromaticity factors with different weights into the original target weights of the image, forming an environmentally adaptive set of adaptation weights. Technically, this step significantly improves the environmental sensitivity of the image selection algorithm, making the constructed standard image database more closely match actual test environment conditions.

[0094] The environmental adaptation weights of each of the evaluation items are sorted, and the standard image databases are determined based on the sorting results.

[0095] Specifically, the environmental adaptation weights of each evaluation item are sorted, and the standard image databases for each item are determined based on the sorting results. This step is the image selection process, aiming to select the most suitable test images for each evaluation item based on their overall adaptation performance. The implementation involves sorting the test images for each evaluation item from highest to lowest according to their environmental adaptation weights, and selecting a number of images based on a preset image quantity threshold or minimum weight threshold to form the standard image database. This process not only reflects the dynamic nature of the weight sorting but also balances the comprehensiveness of the evaluation with data resource control. The resulting standard image database has flexible characteristics that vary depending on the environment, helping to maintain evaluation consistency, reliability, and comparability across various test scenarios, fully leveraging the intelligent adaptive advantages of the MLED evaluation method.

[0096] Preferably, the step of analyzing the suitability of each test image under the current ambient light conditions based on the ambient light intensity to obtain the light intensity suitability factor includes:

[0097] The test environment illuminance is mapped according to the preset illuminance level range to obtain the target level range corresponding to the test environment illuminance.

[0098] Specifically, ambient illuminance refers to the lighting intensity in the current test environment, typically measured in Lux. For example, an office environment is generally 300–500 Lux, while a cinema environment may be below 50 Lux. The system's preset illuminance level ranges can be divided into several illuminance levels, such as low (0–100 Lux), medium (101–500 Lux), and high (501–1000 Lux). The purpose of this step is to classify the currently measured illuminance value into a relatively stable range, facilitating subsequent image adaptability analysis. In this process, the system collects ambient illuminance values ​​in real time using an ambient light sensor, calls a range mapping table, determines which preset range the illuminance value belongs to based on comparison relationships, and outputs the corresponding level label as the target level range. By using range mapping, the interference of minor numerical fluctuations on the evaluation process can be shielded, improving the stability and decision-making efficiency of the system's environmental identification.

[0099] Classify each test image in the standard image database to obtain the image category of each test image;

[0100] Specifically, image categorization involves classifying and labeling test images in a standard image database based on their functional characteristics. Examples include brightness test images, chromaticity test images, contrast test images, and distortion test images. The purpose of this step is to establish an evaluation sensitivity foundation for each type of test image, enabling subsequent personalized analysis of its environmental adaptability. During implementation, the system automatically categorizes images based on their intended use or image tag information, either by calling the image attribute recognition module or by quickly classifying them using embedded tag fields in the image metadata. Through image category recognition and differentiation, the system can construct a multi-dimensional adaptation mechanism based on differences in image function, matching more accurate evaluation models to different categories of images and enhancing the system's targeted intelligent adaptability.

[0101] Based on the preset mapping relationship between image category and light sensitivity level, the target light sensitivity level corresponding to each test image is obtained;

[0102] Specifically, the light sensitivity level is used to characterize the degree of response of different types of images to changes in light intensity. For example, high-contrast images remain highly readable under low light conditions, indicating low light sensitivity; while low-grayscale images are difficult to identify in low light and belong to the high light sensitivity type. The purpose of this step is to quantify and map image categories to their corresponding light adaptability, providing parameter basis for subsequent adaptation factor calculation. In implementation, the system assigns a light sensitivity level to each test image based on a preset image category and light sensitivity level correspondence table (such as "brightness image → low sensitivity", "grayscale image → high sensitivity", etc.) by looking up the table. The technical advantage of this step is that it constructs adaptation rules based on image features, enabling more fine-grained light adaptation adjustment and effectively improving the accuracy of evaluation results for different images under non-ideal lighting conditions. Fluctuation analysis is performed on the light intensity of the test environment to obtain light stability, wherein the light stability includes the light change rate and light fluctuation frequency.

[0103] For the target level range, the target illuminance sensitivity level and the illuminance stability are weighted and fused to obtain the illuminance adaptation factor.

[0104] Specifically, illumination stability is a measure of the time-varying characteristics of illumination intensity in a test environment. Illumination change rate refers to the rate at which illumination values ​​change per unit time, while illumination fluctuation frequency represents the number of times illumination intensity changes per unit time. This step aims to identify whether the illumination environment is stable during the test, thereby determining whether image weights need to be adjusted or the evaluation confidence level reduced. It is typically implemented by the system continuously recording illumination value sequences at a fixed sampling frequency and calculating the average rate of change and zero-crossing frequency of the difference sequences to obtain the illumination stability index. The introduction of illumination stability enhances the system's dynamic response capability, enabling the evaluation process to dynamically optimize the confidence model based on the actual environmental jitter, thereby reducing the risk of evaluation errors caused by environmental instability. The illumination adaptation factor is a measure reflecting the comprehensive adaptation capability of a test image in the current test illumination environment. Its calculation requires simultaneous consideration of the actual illumination level, the image's illumination sensitivity, and the stability of the current environment. The purpose of this step is to generate an adaptation factor that guides image weight correction through a mathematical model under the influence of multiple parameters, achieving a quantitative expression of the matching relationship between the environment and the image. The implementation process includes: converting illumination levels and sensitivity levels into standardized quantitative scores, combining them with illumination stability parameters, and calculating the illumination adaptation factor value for each image in the current environment using a weighting function (such as linear weighting, fuzzy inference, or neural network fusion). This step technically introduces a multi-dimensional cross-evaluation strategy, effectively balancing the dual impact of image characteristics and external environmental changes on evaluation reliability, significantly improving the accuracy of environmental matching and dynamic adjustment capabilities of test image selection.

[0105] Preferably, the step of analyzing the adaptability of each test image under the current environmental chromaticity value based on the test environment chromaticity value to obtain the chromaticity adaptation factor includes:

[0106] The deviation between the test environment chromaticity value and the preset standard chromaticity value is calculated to obtain the chromaticity deviation value;

[0107] Specifically, the test environment chromaticity value refers to the color attribute of the ambient light in the current test environment, usually expressed in CIE1931 chromaticity coordinates (x, y) or CCT (correlated color temperature). For example, the CCT under natural light is approximately 5500K, while under fluorescent light it may appear greenish or bluish. The standard chromaticity value refers to the ambient light chromaticity benchmark recommended or calibrated in the MLED terminal standard test, serving as the ideal chromaticity condition for evaluation activities. The purpose of this step is to identify whether the chromaticity of the light source in the actual test environment deviates from the standard, quantify the color deviation, and provide basic data for subsequent image evaluation and adaptation. Its implementation typically includes: acquiring the (x, y) coordinates or CCT value of the current environment through a chromaticity sensor; retrieving the preset standard chromaticity value; comparing the two according to a defined chromaticity difference function (such as ΔE or coordinate difference), and outputting a numerical chromaticity deviation value. This step, by introducing a chromaticity deviation quantification index, provides accurate input for the system's subsequent image adaptation algorithm, significantly improving the system's ability to perceive environmental color changes and identify chromaticity interference.

[0108] Based on the chromaticity deviation value, determine the direction and level of chromaticity deviation;

[0109] Specifically, the chromaticity deviation direction refers to the deviation trend of the current ambient light chromaticity relative to the standard chromaticity, such as a reddish, blued, or greenish bias. This is typically obtained by comparing the vector directions of the current chromaticity coordinates with those of the standard chromaticity coordinates. The chromaticity deviation level is a graded representation of the chromaticity deviation value. For example, ΔE < 1 is defined as no perceptible difference, ΔE between 1 and 3 is a slight deviation, and ΔE > 5 is a significant deviation. The purpose of this step is to further refine the characteristics of the chromaticity deviation to support subsequent targeted correction or dynamic weighted processing in image evaluation. The implementation process includes: calculating the direction vector of the chromaticity deviation value output in the first step and the standard chromaticity value to identify the color shift trend; and determining the corresponding chromaticity deviation level label by combining it with the established deviation level classification table. This step extracts chromaticity deviation information (direction + level) from two dimensions, enabling the system to perform targeted deviation compensation for color-sensitive images during the evaluation process, thereby effectively suppressing evaluation errors caused by ambient color shift.

[0110] The chromaticity adaptation factor is obtained by weighting the chromaticity deviation direction, the chromaticity deviation level, and the preset chromaticity sensitivity corresponding to each test image.

[0111] Specifically, the chromaticity adaptation factor is a comprehensive index representing the adaptability of a specific test image under the current chromaticity environment. Its calculation requires comprehensive consideration of the direction and severity of the current environment's color shift, as well as the image's sensitivity to chromaticity changes. Chromaticity sensitivity is a scalar measure of the degree of perceptible difference for each type of test image under chromaticity change conditions. For example, color accuracy evaluation images typically have high chromaticity sensitivity, while black and white images have virtually no response to chromaticity changes. The purpose of this step is to form a consistent metric for color adaptability in the image environment adaptation strategy, for use in weight adjustment or evaluation confidence correction. In the implementation process, the system first matches and judges the direction of chromaticity deviation with the sensitive direction of the image content, then sets an adaptation adjustment factor according to the deviation level, performs a weighted fusion operation with the corresponding image's sensitivity parameter, and outputs a continuous value for the chromaticity adaptation factor. This step technically implements an evaluation robustness enhancement mechanism under color interference, maintaining the stability and objectivity of image test evaluation results under dynamic lighting color temperature or environmental color shift conditions.

[0112] Preferably, please refer to Figure 2 The step of evaluating the MLED display terminal based on the preset evaluation strategy corresponding to each of the evaluation items, combined with the real-time image data and the target standard image, to obtain the target evaluation result corresponding to the display quality of the MLED display terminal includes:

[0113] Based on the terminal type and the usage scenario type, determine the evaluation priority corresponding to each evaluation item;

[0114] Specifically, terminal type refers to the specific application category of the MLED display terminal, such as for cinemas, automotive displays, portable devices, or medical displays; usage scenario type refers to the actual application environment of the terminal, such as high-brightness environments, nighttime viewing, outdoor strong light, or dark room demonstrations; evaluation items are dimensions for quantitatively evaluating the performance of MLED displays, such as brightness uniformity, color reproduction, response time, and contrast ratio. The purpose of this step is to construct a targeted performance evaluation priority sequence based on the actual application characteristics of the terminal, avoiding wasting resources on irrelevant dimensions and improving the efficiency and practicality of the evaluation. The implementation process includes: first, the system obtains the terminal type and usage scenario type based on the recognition module or manual input; then, it calls a preset priority matching rule table, which stores the evaluation item weights or priority labels corresponding to each type of terminal and scenario, and finally outputs an evaluation priority list. Through this step, the evaluation order of different test items in different device applications can be dynamically adjusted, ensuring that test resources are concentrated on core indicators, achieving precision and differentiation in the evaluation strategy.

[0115] Based on the evaluation priorities described above, determine the target priorities that meet the preset priority requirements, and obtain the target evaluation items corresponding to the target priorities;

[0116] Specifically, target priority refers to the priority threshold level defined by the current evaluation requirements or system strategy. For example, evaluation items with the highest priority will be considered key evaluation content. Target evaluation items refer to specific test items selected to participate in subsequent evaluation processes under the current priority strategy. The purpose is to further filter core items that meet the set priority conditions from the overall set of evaluation items, thereby controlling the evaluation scope and improving system operating efficiency. Based on the priority of each evaluation item obtained in the first step, matching and filtering are performed to finally select a set of target evaluation items for subsequent strategy matching. Through this step, the most relevant items for the current environment or terminal can be efficiently filtered from a large number of test dimensions, significantly improving system response speed and reducing computational redundancy, thereby improving the decision-making efficiency and controllability of the overall evaluation chain.

[0117] According to the preset evaluation strategy, obtain the target evaluation strategy corresponding to the target evaluation item;

[0118] Specifically, the preset evaluation strategy refers to the set of evaluation rules predefined in the system based on extensive experience in evaluating display terminals. It typically includes test methods, judgment criteria, weighting models, and sample image selection methods for each evaluation item. The target evaluation strategy, on the other hand, is a specific strategy unit selected from the strategy set, corresponding one-to-one with the target evaluation items, and used for actual evaluation execution. The purpose of this step is to assign specific, executable evaluation methods to the selected target evaluation items, establishing a bridge between item definition and specific evaluation actions. This step involves: first, reading the list of target evaluation items; then, searching the strategy definition corresponding to each item in the strategy database and constructing a target evaluation strategy list. This step ensures that the system's evaluation mechanism maintains consistency while enabling personalized selection, supporting subsequent parallel execution or weighted fusion of multiple strategies, thereby achieving a flexible and reusable evaluation system architecture.

[0119] Based on the target evaluation strategy, and combining the real-time image data and the target standard image, the MLED display terminal is evaluated to obtain a first evaluation sub-result;

[0120] Specifically, the target evaluation strategy refers to the evaluation execution method corresponding to the target evaluation item, including specific image comparison algorithms, parameter extraction rules, and evaluation judgment thresholds. Real-time image data is the displayed image obtained by the image acquisition device when displaying the test image corresponding to the target evaluation item on the MLED display terminal, used to characterize the actual output. The target standard image is the original image corresponding to the target evaluation item, used as a reference template for comparison and analysis with the display effect. The purpose of this step is to perform the core image quality detection task, and based on the comparison results between the standard reference image and the real-time acquired image, obtain the quantitative index that best represents the performance of the target evaluation item, i.e., the first evaluation sub-result. In specific implementation, the system displays the standard image to the MLED terminal under test and acquires the displayed image data in real time through the calibration camera; then, it uses image processing algorithms (such as SSIM structural similarity analysis, grayscale offset calculation, color space comparison, etc.) to perform pixel-by-pixel or region feature comparison, extract evaluation indicators, and output sub-results. Through this step, quantitative evaluation of the MLED terminal in key performance dimensions can be achieved, a real-time feedback link can be established, and a reliable evaluation basis can be provided for subsequent decision-making.

[0121] Based on the first evaluation sub-result and the preset evaluation early termination condition, determine whether the first evaluation sub-result meets the evaluation early termination condition;

[0122] Specifically, the early termination condition for evaluation is a pre-defined strategy judgment mechanism used to determine whether the display terminal has significant quality problems based on the evaluation values ​​of key indicators, such as insufficient brightness or severe crosstalk. In this case, continuing to execute other non-critical items will not change the overall conclusion of non-compliance. The purpose of this step is to quickly determine whether there are performance anomalies through the results of a small number of high-priority items, thereby realizing a dynamic termination mechanism for the evaluation process. In specific implementation, the obtained first evaluation sub-result is compared with a predefined early termination threshold, such as whether the standard deviation of brightness uniformity > 20% is valid, or whether the chromaticity deviation ΔE exceeds the acceptable upper limit. If the termination condition is met, it indicates that the current terminal is unqualified in terms of basic performance. The beneficial effect of this step is that it significantly improves evaluation efficiency, avoids resource waste, and is particularly suitable for the need for rapid screening of non-conforming products in production line quality inspection scenarios.

[0123] If the first evaluation sub-result meets the early termination condition of the evaluation, then the first evaluation sub-result will be directly used as the target evaluation result.

[0124] Specifically, the target evaluation result is the final judgment of the overall display quality of the MLED display terminal, which can be a pass / fail label or a comprehensive score with different grades. This step, when it is determined that the terminal clearly fails to meet the requirements of key performance indicators, immediately stops further evaluation of other non-critical items and directly uses the existing first evaluation sub-result to construct the final evaluation result. The purpose of this step is to avoid redundant calculation processes, ensure that evaluation resources are concentrated on high-value items, and shorten the overall evaluation cycle. In specific implementation, after confirming that the first evaluation sub-result has triggered the preset termination condition, its result is marked as "Terminated Valid" and encapsulated as the final output (e.g., marked as "Fail" or "Quality Grade: D"), while skipping the subsequent strategy loading and image analysis processes for non-target items. This design has significant efficiency advantages, supports high-speed screening for large-volume products, and helps build a real-time responsive quality inspection system.

[0125] If the first evaluation sub-result does not meet the early termination condition of the evaluation, then according to the evaluation priority, the evaluation strategy corresponding to the non-target evaluation item in each evaluation item is obtained in sequence.

[0126] Specifically, non-target evaluation items refer to evaluation dimensions that were not initially set as key judgment criteria, such as contrast response, image edge sharpness, and afterimage analysis. Evaluation priority is determined by pre-assigned importance levels for each item based on the type of MLED terminal (e.g., commercial display screen, cinema screen) and usage scenario (e.g., close-range reading, wide-angle viewing), thus determining the evaluation order. Evaluation strategies consist of image processing methods and judgment criteria for different items. The purpose of this step is to improve the completeness and scientific rigor of the evaluation results by introducing supplementary items when key indicators cannot fully reflect the overall quality. In the specific implementation, the remaining non-target items are sorted by priority, and the corresponding evaluation strategies are loaded sequentially. This includes calling matching test image templates, setting specific image analysis algorithms (e.g., edge detection, frequency domain distortion ratio), and preparing for subsequent multi-dimensional fusion. The technical advantage of this step lies in maximizing the evaluation effect with limited resources through priority-driven evaluation path scheduling, meeting the needs of high-reliability quality assessment.

[0127] Based on the evaluation strategy corresponding to the non-target evaluation items, and combined with the real-time image data and the target standard image, the MLED display terminal is evaluated sequentially to obtain each second evaluation sub-result;

[0128] Specifically, the second evaluation sub-result refers to the independent evaluation result of each item obtained after evaluating non-target evaluation items, such as the sharpness degradation rate and gamma response curve fitting degree. The purpose of this step is to conduct a more complete performance analysis of the MLED terminal by acquiring multi-dimensional supplementary indicators when key indicators cannot independently determine the overall display quality. In practice, after retrieving the evaluation strategy for non-target items, the terminal is controlled to display specific test images according to the strategy requirements. Real-time image data is then acquired by the image acquisition system, followed by targeted algorithmic extraction and evaluation calculations to generate a separate sub-result record for each item. During this process, the evaluation algorithm can cover various methods such as histogram equalization analysis, frequency domain distortion estimation, and edge sharpness gradient comparison. This step effectively supplements the performance dimensions not covered by key indicators, enhances the system's fine-grained recognition capability of display quality, and provides necessary data support for the final comprehensive evaluation.

[0129] Based on the first evaluation sub-result and in combination with each of the second evaluation sub-results, the target evaluation result is determined.

[0130] Specifically, the target evaluation result is the final assessment conclusion derived from the comprehensive evaluation sub-results across multiple dimensions, representing the overall display performance level or pass / fail status of the MLED terminal under test in the current testing environment. The process can be implemented based on weighted fusion, logical reduction, or rule-based scoring mechanisms. The core objective of this step is to form an accurate, reliable, and consistent evaluation conclusion, provided that multiple evaluation sub-results are representative. In practice, firstly, weight coefficients are assigned to the first and second evaluation sub-results based on the importance level of each item. Then, fusion methods such as weighted average, fuzzy inference, and interval scoring are executed, combined with preset level classification rules, to output the target evaluation result (e.g., A, B, C, or pass / fail). The beneficial effect of this step is to ensure the comprehensiveness and interpretability of the evaluation, considering not only key indicators but also other important but non-fatal performance indicators, providing an accurate data foundation for terminal grading management, factory quality control, and user experience prediction.

[0131] Preferably, determining whether the first evaluation sub-result meets the early evaluation termination condition based on the first evaluation sub-result and the preset early evaluation termination condition includes:

[0132] The target evaluation items are classified to determine the target evaluation item type, and the parameter threshold set corresponding to the early termination condition of the evaluation for that target item type is obtained.

[0133] Specifically, the target evaluation item refers to the key items selected in the evaluation priority, such as brightness uniformity, color accuracy, or grayscale response. The target evaluation item type is a label that categorizes the above items logically, such as brightness, color, or structure, and is used to match the corresponding termination judgment model. The parameter threshold set refers to the set of numerical boundary conditions preset for each item type for early termination judgment, such as brightness uniformity standard deviation not exceeding 15% and chromaticity deviation ΔE not exceeding 3.0. The purpose of this step is to prepare parameters for subsequent termination judgment, and to achieve scalability and customizability of the termination strategy through item classification. In the specific implementation, the current target evaluation item is first semantically classified and mapped to a predefined type system, and then the associated termination threshold set is indexed according to the type as the judgment standard for subsequent comparison. This step improves the flexibility of strategy management through the three-layer abstraction of item-type-threshold and supports differentiated termination mechanisms for multiple types of terminals.

[0134] The evaluation parameters are extracted from the first evaluation sub-result to obtain a set of key evaluation parameters;

[0135] Specifically, the evaluation parameters are specific technical indicators extracted from the first evaluation sub-result, such as the brightness center value, maximum deviation rate, and average ΔE, which represent the actual performance of the terminal in a certain evaluation item. The key evaluation parameter set contains all evaluation values ​​associated with the termination conditions of the item. The purpose of this step is to parametrically parse the complex structured data contained in the evaluation sub-result so as to compare it with the threshold conditions one by one. In terms of implementation, according to the parameter extraction rules preset for the current target item type, structural indicator information is extracted from the image analysis results to construct a parameter set with comparable and quantifiable characteristics. This process supports parallel extraction of multiple parameters and can automatically adapt to the image processing output format. This step has the advantages of high extraction efficiency and standardized structure, providing a necessary foundation for the next step of normalization and judgment logic.

[0136] Based on the evaluation logic type corresponding to each parameter threshold in the parameter threshold set, type mapping and normalization processing are performed on each evaluation parameter in the key evaluation parameter set to obtain the standard parameter set corresponding to each parameter threshold.

[0137] Specifically, the evaluation logic type specifies the comparison method or importance logic of each parameter in the termination condition, such as smaller is better, absolute difference limit, interval judgment, etc.; normalization processing refers to converting evaluation parameters of different dimensions and magnitudes into a unified standard (such as 0-1 or negative / positive scores) to obtain a set of standard parameters that facilitates logical judgment. The core purpose of this step is to make different parameters logically comparable and have a unified judgment entry point by standardizing parameter representation. In the specific implementation process, the system executes the corresponding normalization strategy for each parameter value according to its logic type. For example, linear stretching is performed on the chromaticity deviation ΔE, and inverse mapping is performed on the luminance standard deviation, mapping the physical value into a unified evaluation index system to form a set of standard parameters. This step improves the cross-parameter versatility of the system, especially in multi-condition combination judgment scenarios, and can effectively support the high-precision comparison logic of the automated decision engine.

[0138] Based on the correspondence between the set of standard parameters and the set of parameter thresholds, each standard parameter value is compared with its corresponding parameter threshold one by one to obtain a set of comparison results for each standard parameter.

[0139] Specifically, the comparison result set refers to the set of Boolean values ​​or status codes (e.g., "pass", "fail", or corresponding flags) after each standard parameter is compared with its corresponding evaluation threshold. Each comparison result reflects whether a certain dimension meets the judgment condition for early termination. The purpose of this step is to establish a mapping relationship between parameter performance and termination strategy, providing clear data support for the final comprehensive judgment. In implementation, the system compares each value in the standard parameter set with the corresponding boundary value in the threshold set according to a one-to-one correspondence logic (e.g., greater than, less than, range inclusive, etc.), and records the comparison result as a Boolean value or quantified deviation level, forming a complete comparison result set. This step enhances the interpretability of the termination judgment and lays the foundation for supporting subsequent fusion judgments, conditional weight adjustments, and other operations, and is suitable for multi-indicator-driven early termination judgment models.

[0140] Based on the set of comparison results, determine whether the first evaluation sub-result meets the early termination condition of the evaluation.

[0141] Specifically, the Boolean states or deviation levels in the comparison result set are logically aggregated to determine whether the overall "early termination condition" is met, thereby deciding whether to terminate the subsequent evaluation process. The early termination condition can be composed of a combination of multiple individual thresholds, such as termination if any one threshold is not met, termination if two or more failures occur, or termination if the weighted score is below a threshold. The purpose of this step is to implement a dynamic process control mechanism based on indicator-based judgment. In the specific implementation, based on the predefined termination judgment logic function under the current target project type, aggregation calculations are performed on the comparison result set, such as AND / OR / multiple threshold judgments, ultimately outputting a global Boolean judgment result to control the evaluation process branches. This step has intelligent discrimination capabilities, enabling "early judgment and early exit" in evaluation decisions, effectively reducing the computational complexity and time consumption of evaluation, and improving the processing capacity of large-scale terminal automated testing.

[0142] Preferably, determining the target evaluation result based on the first evaluation sub-result and in combination with the second evaluation sub-result includes:

[0143] Obtain the first evaluation weight corresponding to the target evaluation item and the second evaluation weight corresponding to each of the non-target evaluation items;

[0144] Specifically, "target evaluation items" refer to key items identified as core in the priority ranking, such as brightness uniformity and color accuracy. Their corresponding "first evaluation weight" reflects the dominant role of this item in the overall evaluation system. "Non-target evaluation items" refer to other secondary but still valuable evaluation items, such as grayscale response speed and crosstalk level. Their corresponding "second evaluation weight" reflects their influence in the overall evaluation. The main purpose of this step is to quantitatively express the influence of each evaluation item by introducing a differentiated weight allocation system, ensuring that the final evaluation results are more objective and application-oriented. In practice, the system can call the corresponding evaluation configuration template or strategy library based on dimensions such as terminal type, application scenario, and customer needs to extract weight values ​​matching the current evaluation task, or dynamically generate weights through a weighted learning algorithm. Through this step, the evaluation system possesses adjustability and structured expression capabilities, supporting the accuracy and scenario adaptability of subsequent result fusion.

[0145] Based on each of the second evaluation weights, the second evaluation results are weighted and calculated to obtain the initial evaluation results;

[0146] Specifically, the "second evaluation result" refers to the set of evaluation scores obtained by performing image analysis and strategy evaluation on each non-target evaluation item, usually in the form of sub-item scores or performance indicators; the "initial evaluation result" is an intermediate summary result formed by weighted summarization of these second evaluation results, reflecting the quantitative contribution of secondary evaluation indicators to the overall quality of the terminal. The purpose of this step is to integrate multiple secondary evaluation items so that they form a holistic reference in the final evaluation system. In the specific implementation process, the system multiplies the score value output by each non-target evaluation item with its corresponding second weight, and then sums or normalizes all products to form a single initial score value. The technical feature of this step is that it achieves a comprehensive expression of sub-item evaluation through a weighted mechanism, avoiding evaluation bias caused by the imbalance of weights between indicators, while enhancing the adjustability and maintainability of the overall evaluation model.

[0147] The initial evaluation result and the first evaluation result are weighted and calculated based on the first evaluation weight and each of the second evaluation weights to obtain the target evaluation result.

[0148] Specifically, the "first evaluation result" is the core result value after evaluating the target evaluation items, reflecting the quality level of the terminal in the main evaluation indicators; the "target evaluation result" is the comprehensive output result used to characterize the overall display performance of the MLED display terminal. The purpose of this step is to integrate the evaluation results of the core evaluation items and secondary items, reflecting subjective focus while taking into account other quality dimensions, achieving a more comprehensive and balanced quality evaluation conclusion. In the implementation process, the system, according to a preset or self-learning adjusted comprehensive weighting strategy, multiplies the first evaluation result by the first evaluation weight, multiplies the initial evaluation result by the second evaluation total weight (or calculates a weighted average of all items), and then weights and merges the two parts to output the final target evaluation value. This method ensures that the priority of core evaluation indicators is not overwhelmed by multiple scores, while enhancing the robustness of the evaluation with the help of non-core items, ultimately achieving high-precision and high-stability automated comprehensive quality judgment.

[0149] Example 2

[0150] Please see Figure 3 This invention provides an MLED evaluation system that integrates a standard image database and dynamic image calibration. The system includes at least one processor, at least one memory, and computer program instructions stored in the memory. When the computer program instructions are executed by the processor, the system implements the method described in Embodiment 1.

[0151] In addition, combined Figure 1The MLED evaluation method that integrates a standard image database and dynamic image calibration described in this embodiment of the invention can be implemented by an MLED evaluation system that integrates a standard image database and dynamic image calibration. Figure 3 A schematic diagram of the hardware structure of the MLED evaluation system that integrates a standard image database and dynamic image calibration provided in an embodiment of the present invention is shown.

[0152] An MLED evaluation system that integrates a standard image database with dynamic image calibration may include a processor and a memory storing computer program instructions.

[0153] Specifically, the processor may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement embodiments of the present invention.

[0154] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0155] Computer-readable media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient media, such as modulated communication signals and carrier waves.

[0156] The processor reads and executes computer program instructions stored in memory to implement any of the MLED evaluation methods that integrate a standard image database and dynamic image calibration in the above embodiments.

[0157] In one example, the MLED evaluation system integrating a standard image database with dynamic image calibration may also include a communication interface and a bus. For example, Figure 3 As shown, the processor 401, memory 402, and communication interface 403 are connected through bus 410 and complete communication with each other.

[0158] The communication interface is mainly used to enable communication between various modules, devices, units and / or equipment in the embodiments of the present invention.

[0159] A bus, including hardware, software, or both, couples together components of an MLED evaluation system that integrates a standard image database with dynamic image calibration. For example, and not limitingly, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, a bus may include one or more buses. While specific buses are described and illustrated in embodiments of the invention, the invention contemplates any suitable bus or interconnect.

[0160] In summary, the embodiments of the present invention provide an MLED evaluation method and system that integrates a standard image database and dynamic image calibration.

[0161] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0162] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0163] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0164] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0165] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0166] It should also be noted that the exemplary embodiments mentioned in this invention describe methods or systems based on a series of steps or apparatus. However, this invention is not limited to the order of the steps described above; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0167] The above description is merely a specific embodiment of the present invention. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the protection scope of the present invention.

Claims

1. A method for MLED evaluation by fusing standard image database and dynamic image calibration, characterized in that, The method comprises: acquiring a standard image database corresponding to a plurality of evaluation items of an MLED display terminal to be evaluated; controlling the MLED display terminal to display a target standard image in the standard image database, and acquiring real-time image data corresponding to the target standard image displayed by the MLED display terminal; evaluating the MLED display terminal according to a preset evaluation strategy corresponding to each evaluation item, combining the real-time image data and the target standard image, to obtain a target evaluation result corresponding to the display quality of the MLED display terminal; The standard image database comprises at least one of the following test images: uniformity test image, linearity test image, distortion test image, brightness test image, chroma test image, display resolution test image and crosstalk test image; The acquisition of the standard image database corresponding to the plurality of evaluation items of the MLED display terminal to be evaluated comprises: acquiring an initial weight of each test image in each evaluation item; acquiring a first weight correction factor corresponding to the terminal type of the MLED display terminal, in combination with a preset mapping relationship between the terminal type and the weight correction factor; acquiring a second weight correction factor corresponding to the use scene type of the MLED display terminal, in combination with a preset mapping relationship between the use scene type and the weight correction factor; correcting the initial weight according to the first weight correction factor and the second weight correction factor to obtain a target weight of each test image in each evaluation item; determining the test image corresponding to each evaluation item as each standard image database according to each target weight; The acquisition of the standard image database corresponding to the plurality of evaluation items of the MLED display terminal to be evaluated comprises: acquiring a test environment illuminance and a test environment chroma value of the MLED display terminal; analyzing the evaluation adaptability of each test image under the current environment illuminance condition according to the test environment illuminance, to obtain an illuminance adaptation factor; analyzing the evaluation adaptability of each test image under the current environment chroma value condition according to the test environment chroma value, to obtain a chroma adaptation factor; weighting each target weight according to the illuminance adaptation factor and the chroma adaptation factor corresponding to each test image, to obtain an environment adaptation weight; sorting each environment adaptation weight in each evaluation item, and determining each standard image database according to the sorting result. 2.The MLED evaluation method of fusing a standard image database and dynamic image calibration according to claim 1, wherein, The analysis of the evaluation adaptability of each test image under the current environment illuminance condition according to the test environment illuminance to obtain an illuminance adaptation factor comprises: mapping the test environment illuminance according to a preset illuminance level interval to obtain a target level interval corresponding to the test environment illuminance; classifying each test image in the standard image database to obtain an image category of each test image; According to a preset mapping relationship between image categories and light sensitivity levels, a target light sensitivity level corresponding to each test image is obtained; The test environment illumination is subjected to fluctuation analysis to obtain illumination stability, wherein the illumination stability includes illumination change rate and illumination fluctuation frequency; The target level interval, the target light sensitivity level and the illumination stability are subjected to weighted fusion calculation to obtain the illumination adaptation factor. 3.The MLED evaluation method of fusing a standard image database and dynamic image calibration according to claim 1, wherein, The analysis of the evaluation adaptation of each test image under the condition of the current environment chroma value according to the test environment chroma value includes: The test environment chroma value and a preset standard chroma value are subjected to deviation calculation to obtain a chroma deviation value; According to the chroma deviation value, a chroma deviation direction and a chroma deviation level are determined; The chroma deviation direction, the chroma deviation level and a preset chroma sensitivity corresponding to each test image are subjected to weighted calculation to obtain the chroma adaptation factor.

4. The MLED evaluation method of fusing a standard image database and dynamic image calibration according to any one of claims 1-3, characterized in that, The evaluation of the MLED display terminal according to a preset evaluation strategy corresponding to each evaluation item, in combination with the real-time image data and the target standard image, to obtain a target evaluation result corresponding to the display quality of the MLED display terminal includes: According to the terminal type and the use scene type, an evaluation priority corresponding to each evaluation item is determined; According to each evaluation priority, a target priority meeting a preset priority requirement is determined, and a target evaluation item corresponding to the target priority is obtained; According to the preset evaluation strategy, a target evaluation strategy corresponding to the target evaluation item is obtained; According to the target evaluation strategy, the MLED display terminal is evaluated in combination with the real-time image data and the target standard image to obtain a first evaluation sub-result; According to the first evaluation sub-result and a preset evaluation early termination condition, it is judged whether the first evaluation sub-result meets the evaluation early termination condition; If the first evaluation sub-result meets the evaluation early termination condition, the first evaluation sub-result is directly taken as the target evaluation result; If the first evaluation sub-result does not meet the evaluation early termination condition, according to each evaluation priority, an evaluation strategy corresponding to a non-target evaluation item in each evaluation item is obtained in sequence; According to the evaluation strategy corresponding to the non-target evaluation item, the MLED display terminal is evaluated in sequence in combination with the real-time image data and the target standard image to obtain each second evaluation sub-result; According to the first evaluation sub-result, in combination with each second evaluation sub-result, the target evaluation result is determined. 5.The MLED evaluation method of fusing a standard image database and dynamic image calibration according to claim 4, wherein, The judgment of whether the first evaluation sub-result meets the evaluation early termination condition according to the first evaluation sub-result and a preset evaluation early termination condition includes: The target evaluation item is classified to determine a target evaluation item type, and a parameter threshold set corresponding to the evaluation early termination condition of the target evaluation item type is obtained; The first evaluation sub-result is subjected to evaluation parameter extraction to obtain a key evaluation parameter set; According to the evaluation logic type corresponding to each parameter threshold in the parameter threshold set, type mapping and normalization processing are performed on each evaluation parameter in the key evaluation parameter set to obtain a standard parameter set corresponding to each parameter threshold; According to the corresponding relationship between the standard parameter set and the parameter threshold set, each standard parameter value and the corresponding parameter threshold are compared one by one to obtain a comparison result set corresponding to each standard parameter; According to the comparison result set, it is judged whether the first evaluation sub-result satisfies the evaluation early termination condition. 6.The MLED evaluation method of fusing a standard image database and dynamic image calibration according to claim 4, wherein, The determination of the target evaluation result according to the first evaluation sub-result and the second evaluation sub-result includes: Obtaining a first evaluation weight corresponding to the target evaluation project and a second evaluation weight corresponding to each non-target evaluation project; According to each second evaluation weight, a weighted calculation is performed on each second evaluation sub-result to obtain an initial evaluation result; According to the first evaluation weight and each second evaluation weight, a weighted calculation is performed on the initial evaluation result and the first evaluation sub-result to obtain the target evaluation result.

7. An MLED evaluation system fusing a standard image database and dynamic image calibration, the system comprising at least one processor, at least one memory, and computer program instructions stored in the memory, which, when executed by the processor, implement the method of any one of claims 1-6.

Images