Game Interface Error Detection Platform Based on Image Recognition Technology

Abstract

The present invention discloses a game interface error detection platform based on image recognition technology, which relates to the field of game detection technology. The platform includes: integrating training samples based on error discrimination elements for the dynamic frame rate of the game; generating an error detection module based on the principle of adversarial training in combination with the training samples; receiving the frame rate image of the game interface, assisting the image recognition module to execute the error discrimination elements for element extraction and element reconstruction to determine the discrimination element matrix; transmitting the discrimination element matrix to the error detection module for discrimination, generating error detection data of the frame rate image and performing a pop-up warning. The present invention solves the technical problems in the prior art that multi-type error discrimination elements in the game interface cannot be effectively processed, and the error detection efficiency and accuracy are low, and achieves the technical effects of accurately detecting errors in the frame rate image of the game interface and effectively performing a pop-up warning, improving the error detection efficiency and accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of game detection, and particularly to a game interface error detection platform based on image recognition technology. Background Art

[0002] With the rapid development of the game industry, the complexity and dynamics of game screens are increasing day by day. In traditional game interface error detection technologies, the discrimination of errors in game screens often lacks comprehensiveness. On the one hand, there are a wide variety of game interface elements, including various icons, texts, colors and other directly presented information, as well as elements such as character behaviors and scene changes deeply integrated with the game scene. Most of the existing technologies can only perform simple detections on single types of elements and cannot accurately handle the situation of multiple types of elements including directly discriminable types and those restricted by the game scene at the same time. On the other hand, the construction of discriminators is not precise enough, lacking a targeted processing mechanism for different nature elements (such as static display elements and dynamic interaction elements in games). Moreover, for elements affected by scene changes, there is no effective method to cope with the discriminant changes brought about by scene transitions, resulting in a significant decrease in the accuracy of error detection during complex game scene switching.

[0003] The prior art has technical problems of being unable to effectively handle multiple types of error discriminant elements in game interfaces and having low error detection efficiency and accuracy. Summary of the Invention

[0004] This application provides a game interface error detection platform based on image recognition technology, which is used to solve the technical problems that the prior art cannot effectively handle multiple types of error discriminant elements in game interfaces and has low error detection efficiency and accuracy.

[0005] In view of the above problems, this application provides a game interface error detection platform based on image recognition technology.

[0006] This application provides a game interface error detection platform based on image recognition technology, and the platform includes:

[0007] Training sample integration module, which is used to integrate training samples for the game dynamic frame rate based on error discrimination elements. The error discrimination elements include a first type of elements for direct discrimination and a second type of elements with game scene constraints; discriminator construction module, which constructs a point discriminator for the first type of elements and a sequence discriminator for the second type of elements based on the adversarial training principle and in combination with the training samples to generate an error detection module. Among them, the sequence discriminator is built-in with a discriminant transfer function that obeys the scene transfer law; discriminant element matrix determination module, which is used to receive the frame rate image of the game interface, assist the image recognition module, perform element extraction and element reconstruction on the error discrimination elements to determine the discriminant element matrix, where the element reconstruction is guided by the detection requirements; comprehensive discrimination module, which is used to transmit the discriminant element matrix to the error detection module for classification discrimination and comprehensive discrimination, generate the error detection data of the frame rate image and perform a pop-up warning.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] Training sample integration module, which is used to integrate training samples for the game dynamic frame rate based on error discrimination elements; discriminator construction module, which constructs a point discriminator for the first type of elements and a sequence discriminator for the second type of elements based on the adversarial training principle and in combination with the training samples to generate an error detection module; discriminant element matrix determination module, which is used to receive the frame rate image of the game interface, assist the image recognition module, perform element extraction and element reconstruction on the error discrimination elements to determine the discriminant element matrix; comprehensive discrimination module, which transmits the discriminant element matrix to the error detection module for classification discrimination and comprehensive discrimination, generates the error detection data of the frame rate image and performs a pop-up warning. It achieves the technical effects of accurately detecting errors in the frame rate image of the game interface and effectively pop-up warning, improving the error detection efficiency and accuracy. Description of the Drawings

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0011] Figure 1 It is a schematic structural diagram of a game interface error detection platform based on image recognition technology provided by an embodiment of this application;

[0012] Figure 2Schematic flow diagram of the discriminant factor matrix determination module in the game interface error detection platform based on image recognition technology provided by the embodiments of the present application.

[0013] Description of reference numerals: Training sample integration module 10, discriminator construction module 20, discriminant factor matrix determination module 30, comprehensive discrimination module 40. Detailed implementation manners

[0014] The present application provides a game interface error detection platform based on image recognition technology, which is used to solve the technical problems that the prior art cannot effectively process multi-type error discriminant factors in the game interface and has low error detection efficiency and accuracy.

[0015] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0016] Embodiment, as Figure 1 shown, the present application provides a game interface error detection platform based on image recognition technology, and the platform includes:

[0017] A training sample integration module 10, which is used to integrate training samples based on error discriminant factors for the game dynamic frame rate, and the error discriminant factors include a first type of factors for direct discrimination and a second type of factors with game scene constraints.

[0018] Specifically, the training sample integration module 10 carefully integrates training samples for the game dynamic frame rate in the game interface error detection platform based on image recognition technology, and this process closely revolves around the error discrimination elements. The error discrimination elements cover a first type of elements for direct discrimination and a second type of elements with game scene constraints. For the first type of elements, they are basic feature elements that can be quickly obtained from the game interface image and used to judge errors. For example, in the game interface, certain specific graphic signs, color combinations, or specific pixel patterns can be classified as the first type of elements. These elements are straightforward and can provide important clues in the initial image analysis. The second type of elements is deeply related to the game scene, and their discrimination requires combining specific game situations and rules. In different game scenes, the manifestation forms and discrimination criteria of the second type of elements are different. For example, in a simulation business game, the construction progress appearance that a building facility should have at a specific development stage and the normal working state of resource collection points (such as the activities of workers, collection animation effects, etc.) belong to the category of the second type of elements. If in an image frame, the displayed construction progress of a building does not match the actual input resources and time, or there are no workers operating at the resource collection point but it shows a normal collection state, these abnormal situations need to be identified based on the second type of elements. The training sample integration module 10 will deeply analyze the dynamic characteristics of the game, accurately locate the ways in which various errors may appear in the frame rate image. For the first type of elements, by widely collecting image fragments in different states during the game operation, including normal interfaces and artificially created or naturally occurring abnormal interfaces, a rich and diverse sample set containing various direct discrimination features is accumulated. For the second type of elements, the module designs and simulates complex operations and event triggers in various game scenes, carefully capturing the image frames corresponding to each key link, covering comprehensive sample data from normal scene processes to various abnormal situations. By rigorously organizing and sorting these samples, a comprehensive, accurate, and highly representative training sample set is provided for the subsequent discriminator construction module, laying a solid data foundation for the entire game interface error detection system, ensuring that the system can accurately identify and judge various game interface errors, and guaranteeing the fluency and stability of the game experience.

[0019] The discriminator construction module 20, the discriminator construction module 20 constructs a point discriminator for the first type of elements and a sequence discriminator for the second type of elements based on the adversarial training principle in combination with the training samples, generating an error detection module, wherein the sequence discriminator is built-in with a discriminant transfer function that obeys the scene transfer law.

[0020] Specifically, the discriminator construction module 20 plays a crucial role in the game interface error detection platform based on image recognition technology. Its operation is based on the advanced adversarial training principle. By deeply analyzing and applying training samples, it constructs a highly targeted point discriminator and sequence discriminator, and then generates an efficient error detection module. The sequence discriminator incorporates a unique discriminant transfer function that follows the scene transition law, greatly enhancing the accuracy and adaptability of error detection. The adversarial training principle provides a solid theoretical framework for the entire construction process. Under this framework, the discriminator and the generator play against each other and jointly optimize. For one type of element, the process of constructing the point discriminator is particularly crucial. As a feature that can be directly obtained from the image, the point discriminator aims to accurately identify whether there are abnormalities in these isolated key feature points. For example, in the game interface, for a specific button icon, the correctness of its shape, color, position and other features can be judged by the point discriminator. By learning a large number of training samples containing the normal and abnormal states of this button, the point discriminator can establish a discriminant model for the button's features, so as to quickly detect any abnormalities in the button in any frame of the image, even if it is just a subtle pixel-level change. Constructing a sequence discriminator for the second type of element requires considering more complex game scene factors. The discrimination of the second type of element often depends on the information presented by a series of consecutive image frames and the logical relationships between them. The sequence discriminator will deeply analyze the change laws of elements in these consecutive frames. For example, in the combat scene of an action game, the continuous action sequence of the character, the change of the front and back frame effects of skill release, etc. It learns the change patterns of elements in normal scenes by learning a large number of samples containing such continuous scenes. The built-in discriminant transfer function follows the scene transition law, which means it can intelligently adjust the discrimination strategy according to the dynamic changes of the game scene. In the game, the scene conversion is frequent and complex, and the discrimination criteria for the same element in different scenes may change. For example, in the ordinary exploration scene of the game, the movement speed of the character may be in a relatively low normal range, but in the pursuit or escape scene, the speed range will change accordingly. The discriminant transfer function can perceive this scene change in real time and apply it to the discrimination process of the second type of element. When the scene changes from one type to another, it will quickly adjust the parameters of the discriminant model according to the pre-learned scene transition probability and the relationship between element changes, ensuring that errors related to the second type of element can still be accurately detected in the new scene. In this way, the discriminator construction module 20 successfully constructs a powerful error detection module, providing reliable technical support for game interface error detection.

[0021] The discriminant factor matrix determination module 30 is used to receive the frame rate image of the game interface, assist the image recognition module, perform element extraction and element reconstruction on the error discriminant factors, and determine the discriminant factor matrix, where the element reconstruction is guided by the detection requirements.

[0022] Specifically, the main responsibility of the discriminant factor matrix determination module 30 is to receive the frame rate image of the game interface and closely assist the image recognition module to perform element extraction and element reconstruction operations based on the error discriminant factors, and finally determine the discriminant factor matrix. Throughout the process, the element reconstruction is always guided by the detection requirements to ensure that the constructed matrix can accurately serve the error detection task. When the frame rate image of the game interface is input into this module, the first thing to do is the element extraction work. Under the guidance of the error discriminant factors, the module will accurately extract various types of information related to error judgment from the image. For the first type of elements that can be directly discriminated, such as obvious features like various icons, text prompts, and color distributions in the game interface, the module can quickly locate and extract their key attributes, such as the contour information of the icon shape, the font and size characteristics of the text, and the pixel distribution of specific color areas. For the second type of elements restricted by the game scenario, the module will deeply analyze complex information such as the character state, action sequence, and scene element layout in the image. Taking the character action sequence as an example, it will extract elements such as the limb position changes, action amplitudes, and time intervals between actions of the character in consecutive frame rate images. These elements are crucial for judging whether the character's behavior conforms to the game logic. After completing the element extraction, it enters the element reconstruction stage. Since different error detection requirements have different focuses, the element reconstruction will reorganize and optimize the extracted elements according to specific requirements. If the detection focus is on the visual effect stability of the game interface, then during the reconstruction process, higher weights will be given to the first type of elements related to the interface layout, color matching, icon display, etc., and they will be arranged and combined in a way that is more conducive to detecting interface visual anomalies to form a more targeted feature vector. When the detection requirement focuses on the correctness of the game character's behavior logic, the second type of elements related to the character's actions and state changes will become the core of the reconstruction. The character action sequence elements in consecutive frame rate images will be integrated to highlight the logical relationship between actions, so as to more effectively judge whether there are errors in the character's behavior. After a series of fine operations of element extraction and reconstruction, the module finally determines the discriminant factor matrix. This matrix will serve as an important input for the subsequent comprehensive discrimination module, providing a key data basis for accurately judging whether there are errors in the game interface and the error types, and strongly promoting the smooth progress of the entire game interface error detection process.

[0023] The comprehensive discrimination module 40 is used to transmit the discrimination factor matrix to the error detection module, perform classification discrimination and comprehensive discrimination, generate error detection data of the frame rate image and issue a pop-up window warning.

[0024] Specifically, the comprehensive discrimination module 40 is responsible for transmitting the discriminant element matrix to the error detection module, and based on this, performs comprehensive and in-depth classification discrimination and comprehensive discrimination work, and finally generates error detection data of the frame rate image and triggers a pop-up window warning to ensure that the game interface errors can be discovered and processed in time. After receiving the discriminant element matrix, it is first transmitted to the error detection module. The error detection module classifies and discriminates various elements in the discriminant element matrix based on the point discriminator and sequence discriminator constructed in the early stage. For a type of element of the direct discriminant class, the point discriminator plays a role and quickly determines whether each isolated element is abnormal. For example, it can accurately detect whether the display of a button in the game interface is normal, including whether its shape, color, position, etc. meet the preset standards. If the button is deformed, the color is abnormal, or the position is offset, the point discriminator can quickly identify and mark it. At the same time, the sequence discriminator discriminates the second type of elements constrained by the game scene. It deeply analyzes the changing rules of the elements in the continuous frame rate image to determine whether it conforms to the normal logic under the game scene. For example, in a battle scene, whether the character's skill release action sequence is correct, from the starting action of the skill to the subsequent special effects display and the impact on the enemy, the sequence discriminator will strictly compare it according to the normal mode learned in advance. After the classification and discrimination are completed, the comprehensive discrimination module performs comprehensive discrimination. It integrates the results from different discriminators, considers the relationship between various elements, and their synergy in the entire game interface and scene. For example, the abnormality of a button may not only be its own display problem, but also related to factors such as surrounding text prompts and scene interactions. The comprehensive discrimination module generates more accurate and comprehensive error detection data by comprehensively analyzing these complex relationships. Once an error is detected, the comprehensive discrimination module immediately activates the pop-up warning mechanism, and the pop-up window will clearly display key information such as the type, location, and possible impact range of the error, so that game developers or testers can quickly locate the problem and take corresponding repair measures. At the same time, the pop-up warning can also be displayed in levels according to the severity of the error, ensuring that urgent and serious errors can be handled first, thereby minimizing the impact of the error on the game experience and ensuring the stable operation of the game.

[0025] In a possible implementation, the discriminator building module further includes:

[0026] Determine a target training scene, wherein the target training scene is any game scene.

[0027] Define a target training scenario, combine training samples based on the secondary elements, and perform adversarial training under a training architecture to generate an initial architecture, where the training architecture consists of a generator and a discriminator.

[0028] Decompose the initial architecture to obtain a sequential discriminator for training.

[0029] Specifically, determining the target training scenario is the starting point of the entire process. This choice is extremely crucial because game scenarios are rich and diverse, and there are significant differences in visual presentations, element associations, and logical rules under different scenarios. Whether it is the fantasy world scenario in a role-playing game, the battle scenario in a competitive game, or the puzzle scenario in a puzzle-solving game, each has its own unique interface elements and dynamic change rules. For example, in the fantasy world scenario of a role-playing game, elements such as the movement of characters, interactions with the environment, and the release of magic skills together constitute a complex picture, and these elements have different manifestations in different areas (such as towns, forests, dungeons, etc.). By determining a specific target training scenario, subsequent training can be focused on the error discrimination elements unique to that scenario, especially the secondary elements closely related to the scenario, thereby improving the pertinence and accuracy of the discriminator.

[0030] The target training scenario is defined, which clarifies the focus scope of subsequent training work. Taking a large-scale 3D action adventure game as an example, assume that the selected target training scenario is the combat scenario in the game. At this time, it is necessary to accurately define various boundary conditions of this combat scenario, including the geographical scope of the scenario, the types and numbers of characters participating in the combat, the time conditions when the combat occurs (such as day or night), and environmental factors that may affect the display of the combat interface (such as weather effects, terrain features, etc.). Through such a definition, the secondary elements closely related to this combat scenario can be effectively isolated and focused, avoiding the interference of other irrelevant scenario factors, and providing a clear and stable context environment for subsequent training. Then, training samples based on secondary elements are introduced. These training samples are carefully selected and preprocessed, specifically for the defined combat scenario. For example, the samples contain consecutive image frames of different characters releasing various skills during combat. These image frames detail a series of dynamic change processes such as the starting actions of skill releases, intermediate special effects, and the final impacts on enemies or the environment. At the same time, it also includes the interface display changes of characters in different states (such as normal combat state, injured state, buff state, etc.), as well as various situations of other elements related to combat (such as changes in health bars, updates of weapon and equipment states, display of combat prompt information, etc.). These training samples not only cover the standard situations under normal combat processes but also include various abnormal situations artificially simulated or collected from actual games, such as incorrect display of skill special effects, delayed updates of character states, incomplete combat prompt information, etc. Then, these rich training samples are incorporated into the training architecture composed of a generator and a discriminator, and the adversarial training process is officially launched. In this process, the main responsibility of the generator is to try to generate realistic image samples according to the given combat scenario characteristics and secondary element patterns. It will use the generative model in deep learning algorithms, the generator part in the generative adversarial network (GAN), to learn the data distribution rules in the training samples and create samples that are highly similar to real combat scenario images in terms of visual effects and logical relationships. The generator will generate an image sequence of a character releasing a specific skill, and this sequence is as close as possible to the situation in the real game in terms of skill special effects, character actions, and interaction effects with the surrounding environment. The discriminator accurately distinguishes between real training samples and samples generated by the generator. The discriminator will conduct a detailed analysis and evaluation of the input image samples based on predefined discrimination criteria. It will consider whether the characteristics of various secondary elements in the image conform to the normal logic of the combat scenario. For example, whether the color, shape, and movement trajectory of skill special effects conform to the game settings, whether the changes in character states are reasonable, and whether the health bar updates are consistent with actual damage calculations. By continuously comparing real samples and generated samples, the discriminator gradually improves its discrimination ability and can more accurately identify the subtle differences between genuine and fake samples. In the interaction process between the generator and the discriminator, they form a dynamic adversarial relationship.The generator continuously tries to deceive the discriminator by generating samples that are more difficult to distinguish; while the discriminator continuously improves its discrimination ability and is not deceived by the generator. As the training continues, both parties continuously optimize their parameters and model structures in this competition. After a large number of iterative trainings, the samples generated by the generator become more and more realistic, and the discrimination accuracy of the discriminator also becomes higher and higher. Eventually, the two reach a relatively balanced state, and at this time, the initial architecture is generated. This initial architecture is the preliminary result of the deep training of the generator and the discriminator in a specific battle scenario. It contains the preliminary recognition and generation ability of the relevant features of the two types of elements in this scenario, laying a solid foundation for subsequent further training and optimization, and thus providing a more targeted and accurate discrimination tool for game interface error detection.

[0031] The initialized architecture model is an overall model architecture including a generator and a discriminator obtained after the previous target training scenario limitation and adversarial training based on two-category element training samples. It includes the layer structure information of the model (such as the type and connection relationship of each layer), parameter information (the number and shape of parameters, etc.), and intermediate data related to the discriminator recorded during the training process (such as the change of the loss function value of the discriminator during the training process, which is used to assist in judging the key parts of the discriminator). Based on the layer structure in the architecture description information, traverse the layers of the model backward from the output layer. For each layer, check its input source and output destination. If the output of a layer mainly flows to subsequent processing parts related to incorrect discrimination (such as the final classification discrimination layer), then mark it as a potentially key layer of the discriminator. At the same time, refer to the intermediate data recorded during the training process. For example, if a certain layer plays a key role in the calculation of the discriminator loss function (such as it is found through gradient calculation that the parameter update of this layer has a great impact on the loss function), then further confirm it as a key layer. Continue to traverse backward until a layer boundary that is clearly distinguishable from the generator input is found, which is determined by checking whether the input data source of the layer comes from the generator part. Once this boundary layer is determined, all layers between the boundary layer and the previously marked key layers are determined as the core part of the discriminator. According to the determined core part of the discriminator, extract the parameters belonging to these layers from the parameter set of the initialized architecture. For each discriminator core layer, according to its parameter structure (such as weight matrix, bias vector, etc.), accurately copy or extract the corresponding parameter values from the overall parameter set. During the extraction process, handle the situation of shared parameters. If it is found that some parameters are shared between the generator and the discriminator, according to the preset rules (such as according to the dominant direction of parameter update during the training process or according to the priority of the current disassembly requirements), decide whether to copy the shared parameters for the discriminator to use independently or to mark them specially for subsequent processing. Reorganize the extracted parameters in the order of the discriminator's layer structure to construct an independent parameter set for the discriminator. This step ensures that the discriminator can correctly load and use these parameters for calculation after being separated from the initialized architecture. According to the layer structure information of the discriminator (obtained from the architecture description information), create a new independent model object using the extracted and reorganized parameter set to represent the sequence discriminator. For each layer, according to its original layer type (such as convolutional layer, fully connected layer, etc.) and connection relationship, use the API of the corresponding deep learning framework (such as TensorFlow or PyTorch) to instantiate the layer object and load the reorganized parameters into these layers. Construct the input layer so that it can receive the two-category element sequence data of the game interface frame rate image, which involves the definition of the input data format and the integration of preprocessing operations. For example, set the shape of the input data to a format suitable for processing sequence data (such as [sequence length, image feature dimension]), and add necessary normalization or encoding operations. Construct the output layer so that it can output discriminant results that meet the requirements of error detection.For example, the output layer is set to a binary classification (normal / abnormal) or multi-classification (corresponding to different error types) layer, and its output can be in the form of class labels or probability values, etc. A small part of the validation data set (including known normal and abnormal binary element sequence data) is used to validate the constructed sequence discriminator model. The validation data is input into the sequence discriminator, and performance metrics such as discriminant accuracy and recall are calculated. If it is found that the performance does not meet the expectations, check for problems such as parameter errors and layer connection errors during the model construction process, and make corresponding adjustments and optimizations. Further fine-tune the model, such as adjusting hyperparameters (such as learning rate, regularization parameter, etc.) or slightly modifying the model structure (such as adding additional hidden layers or adjusting the number of parameters of the layer), to improve the performance of the sequence discriminator. After processing the above algorithm steps, the trained sequence discriminator model is output, and this model can independently perform error discrimination on the binary element sequence in the game interface frame rate image, providing key support for subsequent comprehensive discrimination and error detection.

[0032] In a possible implementation manner, the discriminator construction module further includes:

[0033] Identify the training samples of the binary elements, perform scene clustering processing, and determine multiple scene clusters.

[0034] Traverse the multiple scene clusters to determine the cluster discrimination probability, where the cluster discrimination probability includes the mean value of the element discrimination probabilities within each scene cluster.

[0035] Taking the target training scene as the baseline and based on the cluster discrimination probability, determine the probability transfer relationship under scene transition.

[0036] Use the probability transfer relationship as the discrimination transfer function.

[0037] Specifically, in the game interface error detection system based on image recognition technology, secondary elements need special treatment due to their scene dependence. First, accurately identify the training samples of secondary elements that are closely related to specific game scenes. These samples cover rich information in various game scenes. For example, in role-playing games, the special effects of skill releases and character injury animations in combat scenes, as well as character interaction actions and environmental element changes in non-combat scenes, all belong to secondary elements, and their manifestation forms and attribute characteristics vary in different scenes. Then, perform scene clustering processing, training with a specific scene as the standard. For example, select the main city scene in the game as the reference scene. For a large number of collected training samples of secondary elements, cluster them according to the similarity between samples. Here, the similarity measurement standard is based on the feature differences of secondary elements in different scenes. For example, the interface display elements (including the position of the dialogue box, text color, expression, etc.) when a character is having a conversation in the main city scene and the elements related to character skill releases (skill icons, special effect ranges, damage value displays, etc.) in combat scenes have significantly different feature patterns. Through the clustering algorithm, samples with similar feature patterns are grouped together to determine multiple scene clusters. In this process, the discriminant model trained with the main city scene as the standard needs to compensate the discriminant results based on the transfer probability between scenes when facing other scenes. Because although there are the same secondary elements in different scenes, their attributes may change. For example, the moving speed range of a character in the main city scene is normally within a relatively low value range, but in a chasing scene, the moving speed range will increase significantly. When detecting whether the moving speed of a character is abnormal in the chasing scene, the discriminant standard in the main city scene cannot be simply applied, but the discriminant result needs to be adjusted according to the transfer probability of the moving speed element between the main city scene and the chasing scene. In this way, no matter what the current scene is, unified analysis can be carried out under the selected reference scene (main city scene), and then based on the pre-determined transfer relationship between scenes, the final discriminant result is reasonably adjusted before output, so as to improve the accuracy and adaptability of error detection and ensure the normal operation of the game interface in various complex scenes.

[0038] Traversing multiple scene clusters to determine the cluster discrimination probability is a crucial step. After obtaining multiple scene clusters through scene clustering processing, a detailed analysis is carried out for each scene cluster. For one of the scene clusters, it is necessary to deeply examine all the training samples contained therein, and these training samples are all related to a specific game scene and the corresponding secondary elements. For the secondary elements in each sample, the preset discrimination criteria and algorithms are used for discrimination to obtain the discrimination probability of each element in each sample. For example, in a scene cluster of an action game, for the element of character skill release, if the skill special effects in the sample are displayed completely and meet the expectations, its discrimination probability may be close to 0 (indicating normal); if the skill special effects are missing or abnormal, the discrimination probability may be relatively high (close to 1, indicating abnormal). After obtaining the discrimination probability of each element in each sample, calculate the average value of the discrimination probabilities of the same element in all samples within this scene cluster, which is the average value of the element discrimination probability of this element within this scene cluster. Summarize the average values of the discrimination probabilities of all elements within the scene cluster to obtain the cluster discrimination probability of this scene cluster. This cluster discrimination probability can overall reflect the discrimination characteristics of the secondary elements within this scene cluster, providing an important basis for subsequent adjustment of discrimination results based on scene transfer. For example, in a scene cluster containing multiple combat scene samples, if the average value of the discrimination probability of the character injury animation element is relatively high, it indicates that the possibility of abnormal character injury animation within this scene cluster is relatively large. This information is helpful for more accurately judging the normality of this element under different scenes and the direction and amplitude of the corresponding discrimination result compensation when comparing and transferring with other scenes in the future.

[0039] Taking the target training scenario as the baseline and determining the probability migration relationship under scenario transfer by means of cluster discrimination probability is a crucial task. The target training scenario, as the reference point for the entire analysis, has known feature distributions and discrimination patterns of two types of elements. For example, if the town scenario in the game is selected as the target training scenario, in this scenario, two types of elements such as the movement speed of the character and the interface display when interacting with NPCs have their specific normal state ranges and discrimination criteria. Then, for other scene clusters obtained through clustering, analyze their cluster discrimination probabilities. The cluster discrimination probability of each scene cluster reflects the degree of deviation of the two types of elements within the cluster from the normal state. For instance, in a combat scene cluster, the cluster discrimination probability of the skill release frequency element is relatively high, indicating that there are relatively more abnormal skill release frequencies within this scene cluster. Then, by comparing the differences in cluster discrimination probabilities of various elements between the target training scenario and other scene clusters, determine the probability migration relationship under scenario transfer. If transferring from the town scenario to the combat scenario, comparing the cluster discrimination probabilities of the two regarding the character movement speed element, it is found that the discrimination probability of this element in the combat scenario is significantly higher than that in the town scenario, which indicates that during the scenario transfer, the probability of the character movement speed being abnormal has changed, and this change trend is the probability migration relationship. For each of the two types of elements, such comparative analysis should be carried out to obtain a complete scenario transfer probability migration relationship matrix. This matrix details the change rules of the discrimination probabilities of each two types of elements when transferring from the target training scenario to other scenarios, providing a key basis for accurately discriminating the states of two types of elements in different scenarios subsequently, enabling the model to dynamically adjust the discrimination strategy according to the scene conversion, and improving the accuracy and adaptability of error detection.

[0040] Taking the determined probability migration relationship as the discriminant migration function is a key measure to accurately distinguish two types of elements in the discrimination scenario. The probability migration relationship is obtained by deeply comparing the cluster discrimination probabilities of various elements in the target training scenario and other scenario clusters, and details the change trends and rules of the discrimination probabilities of each two-type element when transferring from the target training scenario to other scenarios. When this probability migration relationship is constructed as the discriminant migration function, it endows the sequence discriminator with the ability to dynamically adapt to different scenarios. For example, in a role-playing game, if the target training scenario is an ordinary map scenario, when entering a special dungeon scenario, the discrimination criteria for two-type elements such as the monster behavior patterns and environmental special effects in the scenario will change. At this time, the discriminant migration function can, based on the previously determined probability migration relationship, automatically adjust the discrimination thresholds and weights for elements such as monster behavior patterns and environmental special effects according to the change of the current scenario from the ordinary map scenario to the special dungeon scenario. If the discrimination probability corresponding to the normal range of the monster attack frequency in the ordinary map scenario is relatively low, and the discrimination probability of this element changes according to the probability migration relationship in the special dungeon scenario, the discriminant migration function can accordingly adjust the discrimination strategy so that the sequence discriminator can still accurately judge whether it is abnormal in the new scenario. In this way, no matter what complex and changeable scenario switching the game is in, the discriminant migration function can utilize the probability migration relationship to adjust the discrimination method for two-type elements in real time, thus greatly improving the discrimination accuracy and adaptability of the entire error detection system for game interface errors in different scenarios, ensuring the stable operation of the game and the good experience of players.

[0041] In a possible implementation manner, the discriminator construction module further includes:

[0042] Based on the error discrimination criteria, determine the element threshold.

[0043] Taking the element threshold as a constraint, divide the positive samples and negative samples, where the positive samples are the detected error detection classes.

[0044] Combined with the training samples of the two-type elements, use the positive samples and the negative samples as the generation targets of the generator, and take the element threshold as the discrimination criteria, and perform alternating iterative training of the generator and the discriminator until convergence to obtain the trained discriminator, where the convergence condition is that the discriminator meets the preset discrimination accuracy.

[0045] Specifically, first, the element threshold is determined based on the error discrimination criteria. The error discrimination criteria are a series of pre-set rules and metrics used to measure whether various elements in the game interface are in a normal state. These criteria cover multiple aspects such as the numerical range of elements, characteristic patterns, and change rules. For example, for the element of the health value display of a game character, its normal numerical range is set between 0 and 100. Then, according to this criterion, an element threshold can be determined, such as 80 (when the health value is lower than 80, it may need attention). For the element of the cooldown time of a character's skill release, there is a reasonable range for its normal cooldown interval, and exceeding this range may be regarded as abnormal, thus determining the corresponding element threshold.

[0046] Next, with the determined element threshold as a constraint, the positive samples and negative samples are divided. Among them, the positive samples are the detected error detection classes, that is, those samples that show violations of the normal range defined by the element threshold in the game interface. For example, in a combat scenario, if the health value of a character suddenly becomes negative or exceeds 100, or the cooldown time of skill release is shorter than the specified minimum value, these samples will be classified as positive samples because they represent possible error situations. The negative samples are those samples that meet the requirements of the element threshold and are in a normal state, such as samples where the character's health value fluctuates within the normal range and the skill cooldown time is normal.

[0047] Then, in combination with the training samples of the two types of elements, the positive samples and negative samples are used as the generation targets of the generator. The generator attempts to generate realistic samples, which include both negative samples that conform to the normal state and positive samples that simulate error situations. In this process, with the element threshold as the discrimination criterion, the alternating iterative training of the generator and the discriminator is performed. The task of the discriminator is to distinguish between real training samples (including positive samples and negative samples) and the samples generated by the generator. In each iteration, the generator adjusts its generation strategy according to the current discrimination result and tries to generate samples that are more difficult to be distinguished by the discriminator; while the discriminator continuously optimizes its discrimination ability based on the new samples and improves the recognition accuracy of positive samples and negative samples. This alternating iterative training continues until the convergence condition is met, that is, the discriminator meets the preset discrimination accuracy. The preset discrimination accuracy is a pre-set target accuracy rate, such as 95% or higher. When the discriminator's accuracy rate reaches or exceeds this preset value when discriminating a large number of verification samples, it is considered that the training has converged. At this time, the obtained trained discriminator already has high accuracy and reliability, and can effectively identify error situations related to the two types of elements in the game interface error detection, providing important technical support for ensuring the normal operation of the game and the good experience of players.

[0048] In a possible implementation manner, as Figure 2 described, the discriminant element matrix determination module further includes:

[0049] A preset error probability is set, wherein the preset error probability limits the consistency of the determination before and after the reconstruction.

[0050] Taking the preset error probability as a constraint and the determination based on the element threshold as a detection requirement, the reconstruction conditions of the error discrimination element are mined.

[0051] Based on the reconstruction condition, the image recognition module is supervised and trained.

[0052] Specifically, in the game interface error detection system, in order to improve the pertinence and efficiency of the discrimination, a series of key operations are carried out around the preset error probability to optimize the training of the image recognition module. First, the preset error probability is set. The setting of this probability value is crucial and provides a key constraint standard for the entire reconstruction and training process. Its main function is to limit the consistency of judgments before and after reconstruction, and to ensure that when processing the erroneous discrimination elements, the judgment logic before and after remains relatively stable and accurate. For example, if the preset error probability is set to a lower value, such as 0.05 (indicating that a 5% error fluctuation is allowed), this means that when performing element reconstruction and subsequent discrimination, the high consistency of the results before and after should be guaranteed as much as possible to avoid misjudgments due to excessive reconstruction or unreasonable adjustments to the discrimination criteria.

[0053] With the preset error probability as a constraint and the judgment based on the element threshold as the detection requirement, the reconstruction conditions of the elements of error judgment are deeply excavated. Taking the dynamic element of the action state in the game task as an example, in the actual judgment, it is found that only judging a certain key position (such as the position of the character's hand action when performing the key operation of the task) can effectively judge whether the task action is normal. Assuming that the task is to carry an object, the key position information such as the contact position between the hand and the object and the angle of the grasping action are important reconstruction conditions. For some elements, only considering its representation in a certain dimension or low dimension can make an accurate judgment. For example, for the lighting effect element in the game scene, only the brightness dimension (low-dimensional representation) can be used to preliminarily judge whether it is abnormal, without the need to fully analyze multiple dimensions such as the color and direction of the light. In this way, under the strict constraint of the preset error probability, the reconstruction conditions that can improve the judgment efficiency without affecting the accuracy are accurately excavated.

[0054] Finally, based on these reconstruction conditions, the image recognition module is supervised and trained. During the training process, the parameters of the image recognition module are continuously adjusted to enable it to better focus on these key reconstruction conditions for discrimination. For example, when training the image recognition module to recognize the action state of a game task, it focuses on the feature extraction and analysis of the key positions of the hands; when judging the lighting effect, it optimizes the sensitivity to the features of the brightness dimension. Through this targeted training, the image recognition module can quickly and effectively identify errors in the game interface while ensuring accuracy, thereby improving the performance of the entire error detection system and providing strong guarantees for the stable operation of the game and the high-quality experience of players.

[0055] In a possible implementation manner, the comprehensive discrimination module further includes:

[0056] Set a detection period, and the detection period is limited based on a preset discrimination time zone.

[0057] Detect and determine the error detection data of the updated single-frame frequency image, and store it in the temporary database.

[0058] Constrained by the detection period, as the error detection data of the updated single-frame frequency image is stored, the error detection data of the invalid frame frequency image is deleted.

[0059] Specifically, in order to manage the error detection data efficiently and orderly, it is necessary to set a detection period, and this detection period is limited based on a preset discrimination time zone. The preset discrimination time zone is a time range determined based on the characteristics of the game, the running frequency, and the comprehensive consideration of the timeliness and accuracy of error detection. For example, for a fast-paced action game, a shorter discrimination time zone is required, such as 0.1 second as a discrimination time zone, because the actions and scenes in the game change rapidly, and errors that affect the game experience may occur in a short time; while for a strategy game, the discrimination time zone can be appropriately extended, such as 1 second as a discrimination time zone, because its operations and interface changes are relatively less frequent.

[0060] In the complex environment where the game is running, meticulous error detection is required for the updated single-frame images to ensure the smoothness and stability of the game experience. When a new single-frame image is generated, the detection process is immediately started. This process covers in-depth analysis of each element in the image, including the first-class elements of the direct discrimination class and the second-class elements with game scene constraints. For the first-class elements, check whether the icons on the interface are complete and clear, whether the colors are normal, whether the shape and position of the buttons are accurate, and whether the text display is garbled or blurred. For example, if the skill icons in the game have missing pixels or distorted colors, the detection system can quickly capture these anomalies. For the second-class elements, the detection will be based on the rules and logic of the game scene. For example, in the battle scene, whether the character's action sequence conforms to the setting of skill release, whether the action feedback of attack and defense is normal, and whether the position relationship and interaction between different characters conform to the game mechanism. Once the comprehensive detection of the updated single-frame image is completed, all detected error information will be integrated into error detection data. These data carry key information such as the type, location, and severity of the error in the frame rate image. Subsequently, this valuable data will be stored in a temporary database, which is specifically used to store these continuously updated single-frame image error detection data, providing data support for subsequent further analysis, processing, and possible error repair prompts, ensuring that game developers can promptly grasp potential problems during the game operation.

[0061] When new single-frame images are continuously generated and error detection is completed, the corresponding error detection data is continuously stored in the temporary database. As this process continues, the timeliness of each data needs to be examined based on the detection cycle, and the error detection data of invalid frame images that are outdated and have lost their current reference value should be deleted in a timely manner. For example, in a game with fast action scenes, the detection cycle is set to a shorter time interval. As the game screen switches quickly, the error detection data of the frame image data of a previous short action moment may no longer be relevant after the new scene and action unfold. If not deleted in time, these invalid data will gradually accumulate, occupy a large amount of storage space, and may also interfere with the subsequent processing and analysis of valid data. For example, in a game with multi-stage tasks, when the player completes a stage and enters the next stage, the error detection data of some frame images in the previous stage is meaningless to the current stage. At this time, based on the detection cycle, these invalid data are automatically identified and cleaned up to ensure that the temporary database always maintains an efficient storage state, so that resources can be concentrated on processing and analyzing valuable error detection data in the current game state, thereby improving the operating efficiency and accuracy of the entire error detection mechanism.

[0062] In a possible implementation, the comprehensive determination module further includes:

[0063] As the temporary database is updated, element error determination is performed on the error detection data within the detection period in different time zones to determine continuous detection data.

[0064] Pop-up warnings are given for the continuous detection data.

[0065] Specifically, the update of the temporary database is a continuous and dynamic process. As new error detection data continuously flows into the temporary database, element error determination is performed on these data within the detection period. This determination process deeply analyzes the error detection data within each specific time interval (i.e., the time zone defined by the detection period). In this time zone, for each element in the error detection data, whether it is a type of element related to the visual presentation elements of the game interface, such as icons, text, colors, etc., or a type of element related to the game scene and character behavior, strict error judgments are made. For example, for an action game, within a specific detection period, check the type-two elements related to the release of character skills to see if the skill special effects are displayed completely and if the action coherence of the skill release conforms to the game settings; at the same time, also check if there are any display abnormalities in the type-one elements such as skill icons on the interface. Through this comprehensive and detailed determination, the error detection data that is related within the same time zone is integrated to determine continuous detection data.

[0066] These continuous detection data represent error information that is coherent and relevant within a specific time interval. When these continuous detection data are determined, a pop-up warning mechanism is triggered. Pop-up warnings are an important part of the interaction between players and game developers, and they present the error information in a clear and prominent way. The pop-up window will list in detail the key contents such as the error types included in the continuous detection data, the occurrence frequency, and the game links that may be affected, so that game developers can quickly locate the problem, and at the same time let players know the abnormalities existing in the current game, so as to take corresponding measures, such as feedback to the developer or waiting for repair, to ensure the stability and smoothness of the game experience.

[0067] In a possible implementation manner, the comprehensive discrimination module further includes:

[0068] Determine the game state of the frame rate image, assign weights to the error discrimination elements, and determine the element weights.

[0069] Based on the element weights, perform weighted calculation on the discrimination output results based on type-one elements and type-two elements to determine the comprehensive discrimination result.

[0070] Set a preset error tolerance interval, perform element error out-of-limit determination on the discrimination output results. If there is an element error out-of-limit or the comprehensive discrimination result is abnormal, give a pop-up warning.

[0071] Specifically, the game state of the frame rate image is determined. This process requires in-depth analysis of the current situation of the game, such as combat scenes, exploration scenes, plot dialogue scenes, or other specific game links. In different game states, the importance of error discrimination elements varies. Based on this understanding, weight assignment operations are performed on the error discrimination elements to determine the element weights. For example, in combat scenes, the weights of secondary elements such as the accuracy of character skill releases and the effectiveness of enemy attacks are relatively high because they directly affect the combat results of the game and the player experience; while the weights of primary elements such as some decorative icons on the interface are relatively low.

[0072] The element weights play a crucial role. Based on the determined element weights, weighted calculations are carried out on the discrimination output results of primary elements and secondary elements. Primary elements involve the directly presented parts of the game interface, such as whether various icons are clear and complete, whether there are errors in text displays, and whether the colors are normal. The discrimination output results of these elements each have their corresponding weights. At the same time, secondary elements are closely related to the game scene, such as whether the character's actions in a specific scene are logical and whether the changes in scene elements follow the set rules. The discrimination output results of these also have corresponding weights. When performing weighted calculations, the influence of each element is measured. For the discrimination result of each primary element, it is multiplied by its corresponding weight to obtain the weighted result. Similarly, the same operation is performed on the discrimination results of secondary elements. Then, all the weighted results of primary elements and secondary elements are summarized, comprehensively considering the importance of each element in the error detection of the entire game interface. After such a rigorous calculation process, the comprehensive discrimination result is finally determined. This result can comprehensively and accurately reflect whether there are errors in the game interface under the current frame rate image and the severity of the errors, providing a reliable basis for subsequent processing.

[0073] Setting a preset error tolerance range is an important part of ensuring the stability and rationality of the system. This preset error tolerance range is carefully determined based on considerations of the impact on the game experience, defining the acceptable error range. After obtaining the discriminant output result, an element error out-of-limit determination is carried out based on this preset error tolerance range. For the output result of each element, it is carefully checked whether it exceeds the set tolerance limit. For example, for the element of the skill cooldown time of a game character, if its normal range is 5 - 10 seconds and the preset error tolerance range is set to float by 1 second up and down, then when the discriminant output result shows that the skill cooldown time is 4 seconds or 11 seconds, it is necessary to further determine whether it exceeds the error range allowed for this element. At the same time, the comprehensive discriminant result should also be concerned. The comprehensive discriminant result is the overall judgment after considering all elements. If this result shows an abnormality, it means that there may be problems affecting the game experience on the overall game interface. Whether an individual element's error exceeds the limit or the comprehensive discriminant result shows an abnormality, the pop-up warning function will be immediately activated. As a key way of communicating with players and developers, the pop-up warning will clearly and prominently display error information, including which specific element has a problem, the degree of error, and the possible impact on the game, etc., so that players can know the game situation in a timely manner and also facilitate developers to quickly locate and solve problems to ensure the normal operation of the game.

[0074] It should be noted that the above-mentioned order of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above describes specific embodiments of this specification. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0075] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the protection scope of the present application.

[0076] This specification and the drawings are only exemplary descriptions of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A game interface error detection platform based on image recognition technology, characterized in that: The platform includes: A training sample integration module, the training sample integration module is used to integrate training samples based on error discrimination factors for the dynamic frame rate of the game, the error discrimination factors include a first type of factors of the direct discrimination class and a second type of factors with game scene constraints, wherein the first type of factors are basic feature elements that are quickly obtained directly from the game interface image and used to judge errors, i.e., static interface elements, and the second type of factors are deeply associated with the game scene, and their judgment needs to be combined with the game context and rules, i.e., the elements whose logic changes in the dynamic scene. In different game scenes, the expression form and judgment criteria of the second type of factors are different; A discriminator construction module, wherein the discriminator construction module is based on the adversarial training principle, combines the training samples, constructs a point discriminator for the first type of elements, constructs a sequence discriminator for the second type of elements, and generates an error detection module, wherein the sequence discriminator has a built-in discriminant transfer function that obeys the scene transfer law; A discriminant factor matrix determination module, the discriminant factor matrix determination module is used to receive a frame rate image of a game interface, assist the image recognition module, execute the error discriminant factor to perform factor extraction and factor reconstruction, and determine a discriminant factor matrix, wherein the factor reconstruction is guided by the detection requirements; A comprehensive discrimination module, the comprehensive discrimination module is used to transmit the discrimination factor matrix to the error detection module, perform classification discrimination and comprehensive discrimination, generate error detection data of the frame rate image and issue a pop-up window warning; Wherein, the discriminator building module also includes: Identify training samples of the two types of elements, perform scene clustering processing, and determine multiple scene clusters; Traversing the plurality of scene clusters to determine a cluster discrimination probability, wherein the cluster discrimination probability of each scene cluster reflects the degree of deviation of the two-class elements in the cluster from a normal state, wherein the cluster discrimination probability includes a mean value of the element discrimination probability in each scene cluster; Taking the target training scene as the baseline and the cluster discrimination probability as the basis, determining the probability migration relationship under the scene transfer, and determining the probability migration relationship under the scene transfer by comparing the cluster discrimination probability differences of the target training scene and other scene clusters in various elements; Based on the probability migration relationship of each second-class element in the scene, a complete scene transition probability migration relationship matrix is ​​obtained; Using the probability transition relationship matrix as the discriminant transition function; Dynamically adjust the discrimination strategy according to the scene conversion. Specifically, automatically adjust the discrimination threshold and weight of each second-class element according to the change from the current scene to the target scene. Wherein, the comprehensive identification module also includes: Determine the game state of the frame rate image, assign weights to the error determination factors, and determine the factor weights; Based on the factor weights, weighted calculation is performed on the discrimination output results based on the first-class factors and the second-class factors to determine a comprehensive discrimination result; A preset error tolerance interval is set, and the element error exceeding the limit is judged on the discrimination output result. If there is an element error exceeding the limit, or the comprehensive discrimination result is abnormal, a pop-up warning will be issued.

2. The game interface error detection platform based on image recognition technology as claimed in claim 1, characterized in that: The discriminator building module also includes: Determine a target training scene, wherein the target training scene is any game scene; Limiting the target training scenario, combining the training samples based on the two types of elements, performing adversarial training under the training architecture, and generating an initialization architecture, wherein the training architecture consists of a generator and a discriminator; The initialization architecture is disassembled to obtain a trained sequence discriminator.

3. The game interface error detection platform based on image recognition technology as claimed in claim 2, characterized in that: The discriminator building module also includes: Determining a factor threshold based on the error discrimination criterion; Using the element threshold as a constraint, dividing positive samples and negative samples, wherein the positive samples are the detected false detection classes; Combined with the training samples of the two types of elements, the positive samples and the negative samples are used as the generation targets of the generator, and the element threshold is used as the discrimination criterion. The alternating iterative training of the generator and the discriminator is performed until convergence to obtain the trained discriminator, wherein the convergence condition is that the discriminator satisfies the preset discrimination accuracy.

4. The game interface error detection platform based on image recognition technology as claimed in claim 3, characterized in that: The discriminant factor matrix determination module also includes: Setting a preset error probability, wherein the preset error probability limits the consistency of determination before and after reconstruction; Taking the preset error probability as a constraint and the determination based on the element threshold as a detection requirement, mining the reconstruction conditions of the error determination element; Based on the reconstruction condition, the image recognition module is supervised and trained.

5. The game interface error detection platform based on image recognition technology as claimed in claim 1, characterized in that: The comprehensive identification module also includes: Setting a detection period, wherein the detection period is limited based on a preset determination time zone; Detect and determine the error detection data of the updated single frame rate image, and store it in a temporary database; With the detection period as a constraint, as the updated error detection data of the single frame rate image is stored, the error detection data of the failed frame rate image is deleted.

6. The game interface error detection platform based on image recognition technology as claimed in claim 5, characterized in that: The comprehensive identification module also includes: As the temporary database is updated, element error determination is performed on the error detection data within the detection period under the time zone to determine continuous detection data; A pop-up window alert is displayed for the continuous detection data.

Images