Game testing method and device, terminal equipment and computer readable storage medium

Abstract

The application discloses a game testing method and device, terminal equipment and a computer readable storage medium. The method comprises the following steps: obtaining a target analysis result corresponding to a target game and picture data output by the target game in a testing process; determining a target probability of the target game based on the picture data and the target analysis result, wherein the target analysis result comprises a rule constraint condition, and the target probability is used for representing a probability that the target game follows the rule constraint condition; and obtaining a testing result of the target game according to the target probability. The method can probabilistically represent the case that the game follows the rule based on the picture data output by the game, so that the testing result is objective, and the automation degree and detection efficiency of the game testing process are improved.

Description

Technical Field

[0001] This application relates to the field of software testing technology, specifically to a game testing method, apparatus, terminal device, and computer-readable storage medium. Background Technology

[0002] Before release, game products typically undergo extensive testing to verify that the game's operation conforms to preset rules and logical constraints. In this process, testers usually determine whether the game has any inconsistencies with the established rules by actually operating the game and observing its visual feedback, based on the game's design principles. However, this method heavily relies on human experience and attention, resulting in low efficiency. Furthermore, after each game rebuild or version update, the entire manual verification process must be repeated, leading to repetitive tasks and significant manpower and time costs. Summary of the Invention

[0003] This application provides a game testing method, apparatus, terminal device, and computer-readable storage medium, which can probabilistically represent the game's adherence to rules based on the game's output screen data, making the test results objective, improving the automation and detection efficiency of the game testing process, and reducing costs.

[0004] The technical solution adopted by this invention to solve the problem is as follows: On the one hand, this application provides a game testing method, including: Obtain the target game's corresponding target analysis results and the screen data output by the target game during the test; Based on the image data and the target analysis results, the target probability of the target game is determined. The target analysis results include the rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints. Based on the target probability, the test results of the target game are obtained.

[0005] In some embodiments of this application, the target probability of a target game is determined based on screen data and target parsing results, including: The image data is input into the first intelligent model, which then identifies the image data and outputs the physical information of the objects in the image data. Based on the object's physical information and the target analysis results, determine the target probability of the target game.

[0006] In some embodiments of this application, the target probability of a target game is determined based on the physical information of the object and the target parsing results, including: The target analysis results and the physical information of the object are input into the second intelligent model, and the initial probability is obtained by reasoning from the second intelligent model. Based on the object's physical information, the initial probability is updated to determine the target probability of the target game.

[0007] In some embodiments of this application, the target parsing result includes at least one rule constraint, the initial probability includes a first probability for each rule constraint, the target probability includes a second probability for each rule constraint, and the initial probability is updated based on the object's physical information to determine the target probability of the target game, including: For any given rule constraint, determine the likelihood probability of that rule constraint based on the object's physical information; Obtain the prior probability of the constraint condition of this rule; Based on the likelihood probability, the prior probability is updated to obtain the posterior probability of the rule constraint. The posterior probability of the rule constraint is weighted and fused with the first probability of the rule constraint to obtain the second probability of the rule constraint.

[0008] In some embodiments of this application, the prior probability is updated based on the likelihood probability to obtain the posterior probability of the rule constraint, including: Based on the prior probability, the third probability of the rule constraint is obtained; Based on the third probability and the likelihood probability, the fourth probability of the rule constraint is obtained; When the fourth probability is greater than the probability threshold, the posterior probability of the rule constraint is obtained based on the likelihood probability, the fourth probability, and the prior probability. When the fourth probability is less than or equal to the probability threshold, the prior probability is determined as the posterior probability of the rule constraint.

[0009] In some embodiments of this application, the target parsing result includes at least one of document parsing result and code parsing result, and the target parsing result is obtained in the following manner: Obtain document information describing the target game; perform semantic analysis on the document information to obtain document parsing results; and / or, Obtain the code data of the target game; parse the code data to obtain the code parsing results.

[0010] In some implementation schemes of this application, the target parsing results include scene entry information and operation space information, and the screen data output during the target game test is obtained in the following way: The third intelligent model automatically controls controllable objects in the target game to perform game trials based on scene entry information and operation space information, so as to achieve automatic testing of the target game. Record the screen data output by the target game during the testing process.

[0011] Secondly, embodiments of the present invention also provide a game testing device, comprising: The acquisition module is used to acquire the target parsing results corresponding to the target game and the screen data output by the target game during the test; The first determining module is used to determine the target probability of the target game based on the screen data and the target analysis results. The target analysis results include rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints. The second determining module is used to obtain the test results of the target game based on the target probability.

[0012] Thirdly, this application also provides a terminal device, which includes: One or more processors; Memory; and One or more applications, wherein the applications are stored in memory and configured to be executed by a processor to implement the game testing method of any of the first aspects.

[0013] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, the computer program being loaded by a processor to execute the steps in the game testing method of any of the first aspects.

[0014] The beneficial effects of this invention are as follows: By obtaining the target analysis results corresponding to the target game and the screen data output by the target game during the test, and combining the screen data with the target analysis results to determine the target probability of the target game following the rule constraints, the test results of the target game are obtained based on the target probability. The game's compliance with the rules can be probabilistically represented based on the screen data output by the game, making the test results objective, improving the automation and detection efficiency of the game testing process, and reducing costs. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of a game testing system provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating one embodiment of the game testing method provided by the present invention; Figure 3This is a schematic diagram of the structure of the dynamic convolutional neural network model provided in the embodiment of the present invention; Figure 4 This is a schematic block diagram of the game testing device provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of an embodiment of the terminal device provided in this invention. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of the stated features.

[0019] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0020] It should be noted that since the method in this application embodiment is executed in a terminal device, the processing objects of each terminal device exist in the form of data or information, such as time, which is essentially time information. It can be understood that if size, quantity, position, etc. are mentioned in subsequent embodiments, they are all corresponding data that exist so that the terminal device can process them. Specific details will not be elaborated here.

[0021] This application provides a game testing method, apparatus, terminal device, and computer-readable storage medium, which will be described in detail below.

[0022] Please see Figure 1 , Figure 1This is a schematic diagram of a game testing system provided in an embodiment of this application. The game testing system may include a terminal device 100, which integrates a game testing device, such as... Figure 1 Terminal devices in the process.

[0023] In this embodiment, the terminal device 100 is mainly used to acquire the target parsing result corresponding to the target game and the screen data output by the target game during the testing process; based on the screen data and the target parsing result, the target probability of the target game is determined, wherein the target parsing result includes rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints; according to the target probability, the test result of the target game is obtained, which can probabilistically characterize the game's compliance with the rules based on the screen data output by the game, making the test result objective, improving the automation level and detection efficiency of the game testing process, and reducing costs.

[0024] In this embodiment, the terminal device 100 can be an independent server, a server network, or a server cluster. For example, the terminal device 100 described in this embodiment includes, but is not limited to, a computer, a network host, a single network server, a set of multiple network servers, or a cloud server composed of multiple servers. The cloud server is composed of a large number of computers or network servers based on cloud computing.

[0025] It is understood that the terminal device 100 used in the embodiments of this application can be a device that includes both receiving and transmitting hardware, that is, a device having receiving and transmitting hardware capable of performing bidirectional communication on a bidirectional communication link. Such a device may include: cellular or other communication devices having a single-line display, a multi-line display, or a cellular or other communication device without a multi-line display. Specifically, the terminal device 100 may be a desktop terminal or a mobile terminal, and the terminal device 100 may also be one of a mobile phone, tablet computer, laptop computer, etc.

[0026] Those skilled in the art will understand that Figure 1 The application environment shown is merely one application scenario of the solution in this application and does not constitute a limitation on the application scenario of the solution in this application. Other application environments may include those that are more specific to this application. Figure 1 The number of more or fewer terminal devices shown, for example Figure 1 Only one terminal device is shown in the image. It is understood that the game testing system may also include one or more other services, which are not specified here.

[0027] In addition, such as Figure 1As shown, the game testing system may also include a memory 200 for storing data, such as target analysis results, screen data, target probability, test results, etc.

[0028] It should be noted that, Figure 1 The schematic diagram of the game testing system shown is merely an example. The game testing system and scenario described in this application are intended to more clearly illustrate the technical solutions of this application and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the evolution of game testing systems and the emergence of new business scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0029] First, this application provides a game testing method. The execution subject of the game testing method is a game testing device, which is applied to a terminal device. The game testing method includes: acquiring the target parsing result corresponding to the target game and the screen data output by the target game during the test; determining the target probability of the target game based on the screen data and the target parsing result, wherein the target parsing result includes rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints; and obtaining the test result of the target game based on the target probability.

[0030] like Figure 2 The diagram shown is a flowchart of an embodiment of the game testing method in this application. The game testing method may include the following steps S201 to S203, as detailed below: Step S201: Obtain the target parsing results corresponding to the target game and the screen data output by the target game during the test.

[0031] In this step, the target game refers to one or more games that need to be tested and evaluated. The target analysis result refers to the structured result formed after analyzing the target game. Specifically, this can be obtained by analyzing the game's documentation or the game's program, mainly including the rules and constraints that the game should follow during operation. These rules and constraints can be limiting requirements set for game behavior, logical relationships, or state changes, used to define the logical standards that the game should meet during normal operation. Screen data refers to the actual visual content output by the game during testing, which can be obtained by recording the game's displayed screen to form video data or image sequence data.

[0032] Step S202: Based on the screen data and the target analysis results, determine the target probability of the target game. The target analysis results include rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints.

[0033] In this step, the target probability is used to characterize the degree to which the target game follows the rule constraints during the test, that is, the degree of consistency between the game's actual performance and the established rules. By using the rule constraints as the judgment standard and the screen data as the objective observation object, the degree of matching between the two is analyzed, thus forming a statistical or probabilistic expression of rule compliance. This transforms the qualitative judgment of "whether it conforms to the rules" based on human experience into a calculable and comparable quantitative result. Introducing probability as a quantitative evaluation result makes rule compliance no longer a simple binary "yes or no" judgment, but rather a continuous quantitative value that reflects different degrees of deviation, improving the precision of the judgment. This allows the computer to identify the difference between minor anomalies and serious rule violations, achieving a more objective and stable logical consistency assessment.

[0034] Step S203: Obtain the test results of the target game based on the target probability.

[0035] In one specific embodiment, after determining the target probability, intermediate data obtained during the calculation of the target probability can be retrieved and comprehensively analyzed to generate test results. For example, when analyzing screen data and rule constraints, the number of times a rule is triggered for verification and the number of times it is judged as violated have already been determined. Then, the data can be judged according to a pre-set judgment logic to identify game defects. For example, the judgment logic can be that when a rule is violated more than a preset threshold during the test, the corresponding rule is judged to have a potential defect, and the relevant screen data range where the number of rule violations exceeds the threshold is automatically marked to form a potential defect record.

[0036] Furthermore, by combining runtime status information during the testing process, contextual analysis can be performed on the marked screen data intervals to determine the game scenario when the defect is triggered. For example, the game's level stage, character state, or interaction behavior when the rules are violated can be analyzed to summarize the triggering conditions for the defect. By comparing multiple violation instances, it is also possible to identify whether there are common scenarios, thereby determining whether the defect is an occasional anomaly or a recurring problem.

[0037] Furthermore, based on the specific rules that have been violated, the actual execution results can be compared and analyzed with the rule constraints to identify the logical deviations and generate recommended modifications. For example, if a rule requires a specific visual effect to be triggered when a certain condition is met, but the effect is not detected in the actual screen, the problem may lie in the condition's judgment logic not being executed correctly, or the relevant triggering mechanism not working. In this case, the recommended modification is to check whether the triggering logic corresponding to the condition is called correctly, or to confirm whether the relevant parameter settings are correct.

[0038] Test results can include a structured test report that integrates the above analysis results. That is, test results can include target probability values, statistics on the number of violations of each rule, specific location markers of potential defects, descriptions of triggering scenarios, and corresponding modification suggestions. Developers can use the test results to specifically test the game program, improving the efficiency of problem fixing in game testing. In one specific implementation, the target probability of the target game is determined based on the screen data and the target parsing result, including: inputting the screen data into a first intelligent model, having the first intelligent model identify the screen data and output the physical information of the objects in the screen data; and determining the target probability of the target game based on the physical information of the objects and the target parsing result.

[0039] In this embodiment, the physical information of moving or changing objects in the image data can be extracted through intelligent recognition, and the game can be judged based on this information to determine whether it follows the rules and constraints. The first intelligent model can be a trained neural network model, specifically a dynamic convolutional neural network structure. Figure 3 As shown, in the feature extraction structure of the dynamic convolutional neural network, a 2×2 dynamic convolution matrix can be set to enable visual recognition of dynamic physical information.

[0040] Specifically, the video data can be input into the first intelligent model, which extracts features from the video frame sequence and outputs structured information related to the objects, including object category, bounding box, trajectory information, and physical attribute estimates. The object can be a moving or changing object in the game scene, a controlled player character, other non-player characters, or other objects in the game world that can change or move, such as falling leaves or moving boxes. The object category refers to the type identification result of the object in the scene, such as different characters or scene elements; the bounding box is used to mark the object's position and range in the scene; the trajectory information reflects the object's motion path in consecutive frames; and the physical attribute estimates are data describing the object's motion state, such as speed range and whether a collision has occurred.

[0041] By extracting the physical information of objects, visual representations can be mapped to measurable physical parameters. These parameters are then matched against the rule constraints in the target analysis results to determine whether the game's execution conforms to the expected logic. When determining the target probability, statistics can be compiled based on the degree of matching between the extracted physical information and the rule constraints. For example, when a rule stipulates that the speed of a certain type of object should be within a specific range, the proportion of actual estimated speeds falling within that range can be statistically analyzed; when a rule requires that a collision must or must not occur under specific conditions, verification can be performed based on collision recognition results. Through a comprehensive judgment of multiple physical attributes, an overall probabilistic evaluation of rule compliance is formed.

[0042] In one specific implementation, the target probability of the target game is determined based on the object's physical information and the target analysis result, including: inputting the target analysis result and the object's physical information into a second intelligent model, and obtaining an initial probability through reasoning by the second intelligent model; updating the initial probability based on the object's physical information, and determining the target probability of the target game.

[0043] In this embodiment, a second intelligent model is introduced for preliminary reasoning to determine the initial probability that the target game meets the rule constraints. Based on this, the initial probability is updated by combining the object's physical information to obtain a more reliable target probability. In this embodiment, the second intelligent model can be understood as a large model with natural language understanding and common-sense reasoning capabilities. This large model can parse the rule constraints expressed in text form in the target analysis results, understand the logical relationships within them, and judge the rationality of a certain behavior by combining general physical common sense or universal logic. For example, when a rule requires an object to decelerate and stop under specific conditions, but physical information shows that it continues to move at high speed, the model can judge based on common sense that this behavior may violate the rule and generate the corresponding initial probability.

[0044] Specifically, the target analysis results and the physical information of the object can be input into the second intelligent model. The second intelligent model can then reason about whether the movement of an object conforms to the design rules or physical logic based on the semantic understanding of the rules and common sense logic, and output the initial probability corresponding to the rule. This provides reasonable prior judgments when the rule description is relatively abstract or there are complex logical relationships.

[0045] After obtaining the initial probabilities, they can be further updated based on the object's physical information. Specifically, a Bayesian update algorithm can be used, which modifies the probabilities based on existing prior probabilities and new observational evidence. In this scheme, the prior probabilities can serve as the initial probabilities, while the object's physical information constitutes the observational evidence. By fusing the physical information with the initial probabilities, an updated probability can be obtained—a more accurate probability value generated after considering actual operational data to assess whether the rules are followed.

[0046] Relying solely on common-sense reasoning from large models may lack sufficient data support when facing complex or boundary situations; relying solely on physical information output by statistical models may overlook the high-level logical intent within the rule text. Therefore, this embodiment first generates initial truth probabilities from a large model, and then uses Bayesian methods to fuse the objective statistical results output by a dynamic convolutional neural network. This combines "rule semantic understanding" with "behavioral data verification," ensuring that probability assessment is both logically sound and data reliable, thereby improving the overall precision and credibility of the test evaluation.

[0047] In one specific implementation, the target parsing result includes at least one rule constraint, the initial probability includes a first probability for each rule constraint, and the target probability includes a second probability for each rule constraint. Based on the object's physical information, the initial probability is updated to determine the target probability of the target game, including: for any rule constraint, determining the likelihood probability of the rule constraint based on the object's physical information; obtaining the prior probability of the rule constraint; updating the prior probability based on the likelihood probability to obtain the posterior probability of the rule constraint; and weightedly fusing the posterior probability of the rule constraint with the first probability of the rule constraint to obtain the second probability of the rule constraint.

[0048] In this embodiment, the target resolution result may include one or more rule constraints. When multiple rule constraints exist, the probability of each rule constraint needs to be calculated and updated separately. Correspondingly, the initial probability includes the first probability of each rule constraint, and the target probability includes the second probability of each rule constraint. Likelihood probability refers to the likelihood of a rule being true given the evidence of observed physical information about the object; prior probability refers to the original probability judgment held regarding the rule's validity before incorporating current observational evidence; posterior probability is the updated result obtained by correcting the rule's validity probability after considering observational evidence; the second probability is the final rule probability formed by fusing the posterior probability and the initial probability, used to characterize the actual degree of compliance of the rule in this test.

[0049] In the specific implementation process, for any rule constraint, the test data can be statistically analyzed based on the physical information of the objects. For example, the number of objects that conform to the rule and the number that do not conform to the rule during the test can be counted to calculate the probability that the rule holds under the current evidence conditions, thus obtaining the likelihood probability. Simultaneously, the overall probability of the evidence, i.e., the total likelihood of observing the current physical information under all possible conditions, also needs to be calculated. Then, the prior probability of the rule is read, and the likelihood probability, the overall probability of the evidence, and the prior probability are substituted into Bayes' theorem for calculation to obtain the posterior probability.

[0050] After obtaining the posterior probability, it does not simply replace the original first probability. Instead, it weights and fuses the posterior probability with the first probability to obtain the second probability of the rule. That is, part of the weight of the final probability comes from the judgment of the large model based on semantic and common sense reasoning, and the other part comes from the data results obtained from the statistics based on actual physical information.

[0051] In one specific implementation, the prior probability is updated based on the likelihood probability to obtain the posterior probability of the rule constraint, including: obtaining the third probability of the rule constraint based on the prior probability; obtaining the fourth probability of the rule constraint based on the third probability and the likelihood probability; when the fourth probability is greater than a probability threshold, obtaining the posterior probability of the rule constraint based on the likelihood probability, the fourth probability, and the prior probability; when the fourth probability is less than or equal to the probability threshold, determining the prior probability as the posterior probability of the rule constraint.

[0052] In this embodiment, the likelihood probability refers to the probability of observing the current evidence given that the rule is true. It can be obtained by statistically analyzing the ratio of the number of samples satisfying the rule to the total number of samples bound by the rule. The prior probability refers to the initial probability of the rule being true before incorporating the current detection data; it can be the initial probability output by the large model. The third probability can be the complementary probability of the prior probability, i.e., the value obtained by subtracting the prior probability from 1. The total probability of evidence refers to the overall probability of observing the current evidence under all possible circumstances. The fourth probability can be the total probability of evidence calculated using the physical information of the object as evidence. The posterior probability is the updated result obtained by correcting the initial probability of the rule after incorporating actual observed evidence (i.e., the physical information of the object). The probability threshold is a preset limit value used to judge whether the calculation result is valid, to avoid unreasonable calculation results when there is insufficient evidence or the denominator is zero.

[0053] Specifically, the physical information of the objects extracted by the dynamic convolutional network is first traversed. The objects constrained by the rules are counted, and the number of samples that satisfy the rules, the number of samples that do not satisfy the rules, and the total number of samples constrained by the rules are counted. When the total number of valid samples is zero, it means that the current data is insufficient to support statistical inference. In this case, the prior probability is directly retained as the posterior probability, thereby avoiding unreliable updates in the absence of evidence.

[0054] When valid samples exist, the likelihood probability can be calculated based on the statistical results. For example, the number of samples that satisfy the rule can be divided by the total number of samples to obtain the probability P(E|T) of observing the evidence when the rule is true. At the same time, the probability P(E|¬T) of observing the evidence when the rule is false can also be calculated. Then, the prior probability P_T corresponding to the rule is read. Specifically, the prior probability can be a preset value or an initial probability generated by a large model.

[0055] Based on this, the total probability of evidence P(E) is further calculated, which is the weighted sum of the probabilities of the rule being true and the rule not being true, to obtain the overall probability value of the current evidence. The formula is: P(E) = P(E|T) × P_T + P(E|¬T) × (1 - P_T). The third probability is 1 - P_T, and the fourth probability is P(E). The total probability of evidence exists as the denominator in Bayes' theorem to ensure that the calculation of the posterior probability satisfies the normalization of the probability distribution.

[0056] Subsequently, when the total probability of the evidence exceeds a preset probability threshold, it indicates that the current statistical data is valid. The posterior probability P(T|E) can then be calculated using Bayes' theorem, which is the product of the likelihood probability and the prior probability divided by the total probability of the evidence. This yields the updated probability that the rule holds under the current evidence conditions. The formula is: P(T|E) = [P(E|T) × P_T] / P(E). If the total probability of the evidence is less than or equal to the probability threshold, it indicates that the statistical data is abnormal or there is a risk of the denominator being zero. In this case, the update process is stopped, and the prior probability is directly used as the posterior probability to avoid division-by-zero errors or extreme values ​​affecting system stability. The probability threshold can be 0.

[0057] In one specific implementation, the target parsing result includes at least one of document parsing result and code parsing result. The target parsing result is obtained by: obtaining document information describing the target game; performing semantic analysis on the document information to obtain document parsing result; and / or, obtaining code data of the target game; and parsing the code data to obtain code parsing result.

[0058] In this embodiment, the document information can be text data describing the target game, including but not limited to game manuals, design documents, rulebooks, etc., which describe the game's rules, logical constraints, and expected behaviors. By performing natural language processing on the document information to understand the meaning, logical relationships, and constraints of the text, unstructured text information can be transformed into structured, computable document parsing results. For example, it can extract the behavioral restrictions, interaction logic, or specific condition triggering rules of different objects in the game. On the other hand, the code data can be the source code or executable logic description of the game program. By parsing the code, the logical constraints, state change rules, and interaction conditions between objects implemented in the program can be directly identified, thereby obtaining code parsing results. Any one or a combination of the document parsing results and the code parsing results can form a complete target parsing result, providing a rule basis for subsequent testing and analysis. Through automated text parsing and code analysis, the virtual world rules corresponding to the target game can be extracted from different levels, establishing a complete, structured, and computable rule set.

[0059] In one specific implementation, the target parsing result includes scene entry information and operation space information. The screen data output during the target game test is obtained in the following way: based on the scene entry information and operation space information, the third intelligent model automatically controls the controllable objects in the target game to perform game trials, so as to realize the automatic testing of the target game; and records the screen data output by the target game during the test.

[0060] In this embodiment, scene entry information refers to the initial state or entry identifier of a specific scene in the game that can be tested or entered, including level number, map coordinates, or trigger conditions. Operation space information refers to the set of actions that a controllable object in the target game can perform, such as the range of actions for character movement, attack, interaction, and item use. The third intelligent model is an artificial intelligence agent used to control game objects. It can automatically generate operation sequences based on scene entry information and operation space information, enabling the controllable object to complete the trial play process in the game world. That is, the intelligent model can issue operation commands to the controllable objects in the game without manual input, simulating player behavior and making the game run according to preset logic. Through the automatic trial play of the artificial intelligence agent, game test data can be generated continuously, stably, and controllably, while reducing human intervention and human error.

[0061] In another embodiment, the testing process of the target game can also be performed without relying on automated artificial intelligence control, but rather according to a pre-designed test script. The test script refers to a pre-defined sequence of operations or scene flow, including the operation order of controllable objects, triggering conditions, interactive actions, and execution timing. During testing, the test script can be executed automatically, controlling game objects to complete operations and recording the game's feedback on these operations, forming a complete video or image sequence data.

[0062] To better implement the game testing method in this application embodiment, based on the game testing method, this application embodiment also provides a game testing device, such as... Figure 4 As shown, the game testing device 400 includes: The acquisition module 410 is used to acquire the target parsing result corresponding to the target game and the screen data output by the target game during the test. The first determining module 420 is used to determine the target probability of the target game based on the screen data and the target analysis result, wherein the target analysis result includes rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints. The second determining module 430 is used to obtain the test results of the target game based on the target probability.

[0063] In this embodiment, by obtaining the target analysis result corresponding to the target game and the screen data output by the target game during the test, and combining the screen data and the target analysis result, the target probability of the target game following the rule constraints is determined. Based on the target probability, the test result of the target game is obtained. The game's compliance with the rules can be probabilistically represented based on the screen data output by the game, making the test result objective, improving the automation and detection efficiency of the game testing process, and reducing costs.

[0064] In some embodiments of this application, the first determining module 420 determines the target probability of the target game based on the screen data and the target parsing result, including: The image data is input into the first intelligent model, which then identifies the image data and outputs the physical information of the objects in the image data. Based on the object's physical information and the target analysis results, determine the target probability of the target game.

[0065] In some embodiments of this application, the first determining module 420 determines the target probability of the target game based on the physical information of the object and the target parsing result, including: The target analysis results and the physical information of the object are input into the second intelligent model, and the initial probability is obtained by reasoning from the second intelligent model. Based on the object's physical information, the initial probability is updated to determine the target probability of the target game.

[0066] In some embodiments of this application, the target parsing result includes at least one rule constraint, the initial probability includes a first probability for each rule constraint, and the target probability includes a second probability for each rule constraint. The first determining module 420 updates the initial probability based on the object's physical information to determine the target probability of the target game, including: For any given rule constraint, determine the likelihood probability of that rule constraint based on the object's physical information; Obtain the prior probability of the constraint condition of this rule; Based on the likelihood probability, the prior probability is updated to obtain the posterior probability of the rule constraint. The posterior probability of the rule constraint is weighted and fused with the first probability of the rule constraint to obtain the second probability of the rule constraint.

[0067] In some embodiments of this application, the first determining module 420 updates the prior probability based on the likelihood probability to obtain the posterior probability of the rule constraint, including: Based on the prior probability, the third probability of the rule constraint is obtained; Based on the third probability and the likelihood probability, the fourth probability of the rule constraint is obtained; When the fourth probability is greater than the probability threshold, the posterior probability of the rule constraint is obtained based on the likelihood probability, the fourth probability, and the prior probability. When the fourth probability is less than or equal to the probability threshold, the prior probability is determined as the posterior probability of the rule constraint.

[0068] In some embodiments of this application, the target parsing result includes at least one of document parsing result and code parsing result. The acquisition module 410 obtains the target parsing result in the following manner: Obtain document information describing the target game; perform semantic analysis on the document information to obtain document parsing results; and / or, Obtain the code data of the target game; parse the code data to obtain the code parsing results.

[0069] In some embodiments of this application, the target parsing result includes scene entry information and operation space information. The acquisition module 410 obtains the screen data output during the target game test in the following manner: The third intelligent model automatically controls controllable objects in the target game to perform game trials based on scene entry information and operation space information, so as to achieve automatic testing of the target game. Record the screen data output by the target game during the testing process.

[0070] This application embodiment also provides a terminal device that integrates any of the game testing devices provided in this application embodiment. The terminal device includes: One or more processors; Memory; and One or more applications, wherein the applications are stored in memory and configured to be executed by a processor from the steps of the game testing method in any of the embodiments described above.

[0071] This application also provides a terminal device that integrates any of the game testing devices provided in this application. For example... Figure 5 As shown, it illustrates a structural schematic diagram of the terminal device involved in the embodiments of this application. Specifically: The terminal device may include components such as a processor 501 with one or more processing cores, a memory 502 with one or more computer-readable storage media, a power supply 503, and an input unit 504. Those skilled in the art will understand that... Figure 5 The terminal device structure shown does not constitute a limitation on the terminal device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein: The processor 501 is the control center of the terminal device. It connects various parts of the terminal device via various interfaces and lines, and performs various functions and processes data by running or executing software programs and / or modules stored in the memory 502, and by calling data stored in the memory 502, thereby providing overall monitoring of the terminal device. Optionally, the processor 501 may include one or more processing cores; preferably, the processor 501 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 501.

[0072] The memory 502 can be used to store software programs and modules. The processor 501 executes various functional applications and data processing by running the software programs and modules stored in the memory 502. The memory 502 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory 502 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 502 may also include a memory controller to provide the processor 501 with access to the memory 502.

[0073] The terminal device also includes a power supply 503 that supplies power to the various components. Preferably, the power supply 503 can be logically connected to the processor 501 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 503 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0074] The terminal device may also include an input unit 504, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0075] Although not shown, the terminal device may also include a display unit, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 501 in the terminal device loads the executable files corresponding to the processes of one or more applications into the memory 502 according to the following instructions, and the processor 501 runs the applications stored in the memory 502 to realize various functions, as follows: Obtain the target game's corresponding target analysis results and the screen data output by the target game during the test; Based on the image data and the target analysis results, the target probability of the target game is determined. The target analysis results include the rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints. Based on the target probability, the test results of the target game are obtained.

[0076] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0077] Therefore, embodiments of this application provide a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), a disk, or an optical disk, etc. A computer program is stored thereon, and the computer program is loaded by a processor to execute the steps in any of the game testing methods provided in embodiments of this application. For example, the computer program loaded by the processor can execute the following steps: Obtain the target game's corresponding target analysis results and the screen data output by the target game during the test; Based on the image data and the target analysis results, the target probability of the target game is determined. The target analysis results include the rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints. Based on the target probability, the test results of the target game are obtained.

[0078] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the detailed descriptions of other embodiments above, which will not be repeated here.

[0079] In practice, each of the above units or structures can be implemented as an independent entity or can be arbitrarily combined to be implemented as the same or several entities. For the specific implementation of each of the above units or structures, please refer to the previous method embodiments, which will not be repeated here.

[0080] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0081] The above provides a detailed description of a game testing method, apparatus, terminal device, and computer-readable storage medium provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A game testing method, characterized in that, include: Obtain the target analysis results corresponding to the target game and the screen data output by the target game during the test; Based on the image data and the target analysis result, the target probability of the target game is determined, wherein the target analysis result includes rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints; Based on the target probability, the test results of the target game are obtained.

2. The game testing method according to claim 1, characterized in that, Determining the target probability of the target game based on the image data and the target parsing result includes: The image data is input into the first intelligent model, which then identifies the image data and outputs the physical information of the objects in the image data. Based on the physical information of the object and the target parsing result, the target probability of the target game is determined.

3. The game testing method according to claim 2, characterized in that, The step of determining the target probability of the target game based on the physical information of the object and the target parsing result includes: The target analysis result and the physical information of the object are input into the second intelligent model, and the initial probability is obtained by reasoning from the second intelligent model. Based on the physical information of the object, the initial probability is updated to determine the target probability of the target game.

4. The game testing method according to claim 3, characterized in that, The target parsing result includes at least one rule constraint, the initial probability includes a first probability for each rule constraint, and the target probability includes a second probability for each rule constraint. Updating the initial probability based on the object's physical information to determine the target probability of the target game includes: For any of the aforementioned rule constraints, the likelihood probability of the rule constraint is determined based on the physical information of the object. Obtain the prior probability of the constraint condition of this rule; Based on the likelihood probability, the prior probability is updated to obtain the posterior probability of the rule constraint. The posterior probability of the rule constraint is weighted and fused with the first probability of the rule constraint to obtain the second probability of the rule constraint.

5. The game testing method according to claim 4, characterized in that, The step of updating the prior probability based on the likelihood probability to obtain the posterior probability of the rule constraint includes: Based on the prior probability, the third probability of the rule constraint is obtained; Based on the third probability and the likelihood probability, the fourth probability of the rule constraint is obtained; When the fourth probability is greater than the probability threshold, the posterior probability of the rule constraint is obtained based on the likelihood probability, the fourth probability, and the prior probability. When the fourth probability is less than or equal to the probability threshold, the prior probability is determined as the posterior probability of the rule constraint.

6. The game testing method according to claim 1, characterized in that, The target parsing result includes at least one of document parsing result and code parsing result, and the target parsing result is obtained in the following way: Obtain document information describing the target game; perform semantic analysis on the document information to obtain the document parsing result; and / or, Obtain the code data of the target game; The code data is parsed to obtain the code parsing result.

7. The game testing method according to claim 1, characterized in that, The target parsing results include scene entry information and operation space information. The screen data output during the target game test is obtained in the following way: Based on the scene entry information and the operation space information, the third intelligent model automatically controls controllable objects in the target game to perform game trials, thereby achieving automatic testing of the target game; Record the screen data output by the target game during the test.

8. A game testing device, characterized in that, include: The acquisition module is used to acquire the target parsing result corresponding to the target game and the screen data output by the target game during the test; The first determining module is used to determine the target probability of the target game based on the screen data and the target parsing result, wherein the target parsing result includes rule constraints, and the target probability is used to characterize the probability that the target game follows the rule constraints; The second determining module is used to obtain the test results of the target game based on the target probability.

9. A terminal device, characterized in that, The terminal device includes: one or more processors, a memory, and one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the processor to implement the game testing method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program, which is loaded by a processor to perform the steps of the game testing method according to any one of claims 1 to 7.

Images