A method of image processing and related apparatus

Abstract

Embodiments of the present application disclose a method for image processing and a related device, at least related to machine learning and other technologies in artificial intelligence. The method only intercepts the cheating interface of an abnormal virtual player in a virtual game when the abnormal virtual player appears in an abnormal game situation in a complete game interface, which not only can intercept an accurate target screenshot, but also can filter out redundant game background information, thereby saving transmission bandwidth and storage resources. The above method comprises: obtaining code information sent by a server, the code information being used to indicate a logical code corresponding to a behavior state of an abnormal virtual player in a running process of a virtual game; determining behavior state information of the abnormal virtual player in the virtual game according to the code information; determining a screenshot strategy according to the behavior state information and the code information; and a terminal device intercepts a pixel region meeting a preset image resolution from a game interface of the virtual game according to the screenshot strategy to obtain a target screenshot.

Description

Technical Field

[0001] This application relates to the field of computer technology, specifically to an image processing method and related apparatus. Background Technology

[0002] When virtual players use cheating methods to manipulate the game, their game interface displays program calculations that shouldn't be shown, such as the location of enemy virtual players. Therefore, using cheating methods severely impacts game fairness, degrades the gaming experience for legitimate players, and generates a large amount of redundant data for the server, seriously affecting the normal operation of the game.

[0003] In traditional anti-cheat solutions, the server typically sends a screenshot command to the terminal device corresponding to the suspicious behavior of a virtual player. The terminal device executes the command to capture the entire game interface and then sends this captured interface back to the server for processing, such as image recognition. In other words, the server in these solutions is unaware of the virtual player's behavior, making the timing of the screenshot entirely dependent on the server's command. This can easily lead to a mismatch between the timing of the screenshot and the virtual player's cheating behavior, preventing the terminal device from taking a screenshot at the appropriate time. Furthermore, the virtual player's cheating behavior usually only occupies a small portion of the complete game screen, while traditional screenshot methods require capturing the entire screen. This results in a large amount of redundant information in the captured game screen, wasting significant transmission bandwidth and storage resources. Summary of the Invention

[0004] This application provides an image processing method and related apparatus that captures the cheating interface of an abnormal virtual player in a virtual game from the complete game interface. This not only captures accurate target screenshots but also filters out redundant game background information, saving transmission bandwidth and storage resources.

[0005] In a first aspect, embodiments of this application provide an image processing method. The method includes: a terminal device acquiring code information sent by a server, the code information being used to indicate the logic code corresponding to the behavioral state of an abnormal virtual player during the operation of a virtual game; the terminal device determining behavioral state information of the abnormal virtual player in the virtual game based on the code information; the terminal device determining a screenshot strategy based on the behavioral state information and the code information; and the terminal device, according to the screenshot strategy, extracting a pixel area satisfying a preset image resolution from the game interface of the virtual game to obtain a target screenshot, the target screenshot being used to reflect the abnormal game situation of the abnormal virtual player in the virtual game.

[0006] Secondly, embodiments of this application provide a terminal device. The terminal device includes an acquisition unit and a processing unit. Exemplarily, the terminal device further includes a sending unit. The acquisition unit is configured to acquire code information sent by a server, the code information being used to indicate the logic code corresponding to the behavioral state of an abnormal virtual player during the operation of a virtual game. The processing unit is configured to: determine the behavioral state information of the abnormal virtual player in the virtual game based on the code information; determine a screenshot strategy based on the behavioral state information and the code information; and, according to the screenshot strategy, extract a pixel area satisfying a preset image resolution from the game interface of the virtual game to obtain a target screenshot, the target screenshot being used to reflect the abnormal game situation of the abnormal virtual player in the virtual game.

[0007] In some optional implementations, the processing unit is configured to: determine a screenshot time point based on the behavior state information, and determine a first position based on the code information, wherein the first position is the screenshot position when the target screenshot is captured from the game interface based on the screenshot time point; determine the screenshot time point and the first position as a screenshot strategy; and, based on the screenshot time point and with the first position as the screenshot center, capture a pixel area that meets a preset image resolution from the game interface of the virtual game to obtain the target screenshot.

[0008] In some alternative implementations, the processing unit is configured to: determine the game gains of the abnormal virtual player in the virtual game based on the behavioral state information; determine the time point corresponding to the maximum gain among the game gains; and determine the time point corresponding to the maximum gain as the screenshot time point.

[0009] In some alternative implementations, the acquisition unit is configured to: acquire the location coordinates of each enemy virtual player in the virtual game, and the location coordinates of abnormal virtual players, based on the code information. The processing unit is configured to: determine the suspiciousness level of each enemy virtual player based on the location coordinates of the abnormal virtual players and the location coordinates of each enemy virtual player; determine the target suspiciousness level from the suspiciousness levels of each enemy virtual player; and the terminal device determines the location coordinates of the enemy virtual player corresponding to the target suspiciousness level as the first location.

[0010] In some alternative implementations, the processing unit is configured to: calculate the coordinate difference information between the location coordinates of the abnormal virtual player and the location coordinates of each enemy virtual player; prioritize each coordinate difference information to obtain priority information, which is used to indicate the priority of the location coordinates of each enemy virtual player; and determine the degree of suspicion of each enemy virtual player based on the priority information.

[0011] In some alternative implementations, the processing unit is configured to: determine a second position based on the code information, the second position being the screenshot position when capturing the target screenshot from the game interface; determine the second position as a screenshot strategy; and, with the second position as the screenshot center, capture a pixel area from the game interface of the virtual game that meets a preset image resolution to obtain the target screenshot.

[0012] In some alternative embodiments, the acquisition unit is used to: acquire the location coordinates of the abnormal virtual player based on the code information. The processing unit is used to determine the location coordinates of the abnormal virtual player as the second location.

[0013] In some alternative implementations, the sending unit is configured to: after capturing a pixel area that meets a preset image resolution from the game interface of the virtual game according to the screenshot strategy to obtain a target screenshot, send the pixel information corresponding to the target screenshot to the server, wherein the pixel information is used by the server to identify and process the target screenshot.

[0014] In some alternative implementations, the sending unit is configured to: send a first message to the server, the first message carrying pixel information corresponding to the target screenshot.

[0015] Thirdly, embodiments of this application provide a server. The server may include a sending unit. Exemplarily, the server may further include a processing unit and an acquisition unit. The sending unit is configured to send code information to a terminal device. The code information is used to indicate the logic code corresponding to the behavioral state of an abnormal virtual player during the operation of a virtual game. The code information is used to instruct the terminal device to determine the behavioral state information of the abnormal virtual player in the virtual game. The behavioral state information is used to instruct the terminal device to determine a screenshot strategy. The screenshot strategy is used to instruct the terminal device to capture a pixel area from the game interface of the virtual game that meets a preset image resolution to obtain a target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game.

[0016] In some embodiments, the acquisition unit is further configured to receive pixel information corresponding to the target screenshot sent by the terminal device after sending code information to the terminal device. The processing unit is configured to perform identification processing on the target screenshot based on the pixel information.

[0017] In some alternative embodiments, the acquisition unit is further configured to acquire a first screenshot sample and a second screenshot sample before performing recognition processing on the target screenshot based on the pixel information. The first screenshot sample is a screenshot taken from the game interface by a first virtual player when the abnormal game situation occurs in the virtual game, satisfying the preset image resolution. The second screenshot sample is a screenshot taken from the game interface by a second virtual player when the abnormal game situation does not occur in the virtual game, satisfying the preset image resolution. The processing unit is further configured to: iteratively train an initial recognition model based on the first screenshot sample and the second screenshot sample to obtain a target recognition model; and process the pixel information based on the target recognition model to perform recognition processing on the target screenshot to obtain the category of the target screenshot.

[0018] In some alternative implementations, the processing unit is configured to: determine cheating features in a target screenshot sample, wherein the target screenshot sample is a screenshot taken from the game interface by the first virtual player when the abnormal game situation occurs in the virtual game, and the screenshot satisfies the preset image resolution; and superimpose the cheating features onto the second screenshot sample to obtain the first screenshot sample.

[0019] In some alternative implementations, the acquisition unit is configured to receive a first message sent by the terminal device, the first message carrying pixel information corresponding to the target screenshot.

[0020] A fourth aspect of this application provides an image processing apparatus, including: a memory, an input / output (I / O) interface, and a processor. The memory stores program instructions. The processor executes the program instructions in the memory to perform the image processing method corresponding to the embodiment of the first aspect described above.

[0021] The fifth aspect of this application provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method corresponding to the embodiments of the first aspect described above.

[0022] The sixth aspect of this application provides a computer program product containing instructions that, when run on a computer or processor, causes the computer or processor to execute the method described above for performing the implementation of the first aspect.

[0023] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0024] In this embodiment, since the code information reflects the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game, after the server sends the code information to the terminal device, the terminal device can determine the behavior state information of the abnormal virtual player in the virtual game based on the code information. Thus, the terminal device can determine a screenshot strategy based on the behavior state information and the code information, and then, under the guidance of the screenshot strategy, capture a pixel area from the game interface of the virtual game that meets a preset image resolution, thereby obtaining a target screenshot that can reflect the abnormal game situation of the abnormal virtual player in the virtual game. Through the above method, the terminal device can fully consider the behavior state information of the abnormal virtual player in the virtual game, so that the screenshot strategy determined by the behavior state information and the code information can match the cheating behavior of the abnormal virtual player. Therefore, by performing the screenshot operation using a screenshot strategy that matches the cheating behavior of the abnormal virtual player, an accurate target screenshot can be captured. Moreover, by capturing a pixel area from the game interface that meets a preset image resolution, the captured target screenshot can reflect the abnormal game situation of the abnormal virtual player in the virtual game. In other words, the target screenshot only occupies a small part of the entire game interface. It only captures the cheating interface of the abnormal virtual player when abnormal game situations occur in the virtual game from the complete game screenshot, thereby filtering out redundant game background information and saving transmission bandwidth and storage resources. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application 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 of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This diagram illustrates game screenshots provided by existing solutions;

[0027] Figure 2 This diagram illustrates a mismatch in screenshot results when taking screenshots based on existing methods.

[0028] Figure 3 A schematic diagram illustrating the creation of complete game screenshots based on existing methods is shown;

[0029] Figure 4 A schematic diagram of the system framework provided in an embodiment of this application is shown;

[0030] Figure 5 A flowchart of an image processing method provided in an embodiment of this application is shown;

[0031] Figure 6 A schematic diagram showing the trigger screenshot provided in an embodiment of this application is illustrated;

[0032] Figure 7 Another flowchart of the image processing method provided in an embodiment of this application is shown;

[0033] Figures 8A to 8C A schematic diagram showing the distribution of cheating features provided in the embodiments of this application is shown;

[0034] Figure 9 A schematic diagram illustrating the construction method of the training samples provided in the embodiments of this application is shown;

[0035] Figure 10 This illustration shows a schematic diagram of constructing a black sample according to an embodiment of this application;

[0036] Figure 11 A schematic diagram illustrating the filtering of anomalous samples based on a GAN model is shown.

[0037] Figure 12 The flowchart of sample expansion and purification provided in the embodiments of this application is shown;

[0038] Figure 13 A schematic diagram of the sample purification process provided in an embodiment of this application is shown;

[0039] Figure 14This paper shows a schematic diagram of the structure of the Mobilenet-v3 large model provided in an embodiment of this application;

[0040] Figure 15 A schematic diagram illustrating the differences in screenshots provided in the embodiments of this application is shown;

[0041] Figure 16 A schematic diagram of the structure of the terminal device provided in an embodiment of this application is shown;

[0042] Figure 17 A schematic diagram of the server structure provided in an embodiment of this application is shown;

[0043] Figure 18 A schematic diagram of the hardware structure of the image processing apparatus provided in an embodiment of this application is shown. Detailed Implementation

[0044] This application provides an image processing method and related apparatus that captures the cheating interface of an abnormal virtual player in a virtual game from the complete game interface. This not only captures accurate target screenshots but also filters out redundant game background information, saving transmission bandwidth and storage resources.

[0045] It is understood that in the specific implementation of this application, user information, user personal data and other related data are involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0046] 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.

[0047] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0048] The image processing method provided in this application is based on artificial intelligence (AI). Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results. In other words, artificial intelligence is a comprehensive technology within computer science that attempts to understand the essence of intelligence and produce a new type of intelligent machine that can react in a way similar to human intelligence. Artificial intelligence also studies the design principles and implementation methods of various intelligent machines, enabling them to possess perception, reasoning, and decision-making functions.

[0049] Artificial intelligence (AI) is a comprehensive discipline encompassing a wide range of fields, including both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies primarily include computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0050] In the embodiments of this application, the main artificial intelligence technologies involved include the aforementioned machine learning and other related fields. For example, it may involve deep learning in machine learning (ML), including artificial neural networks, etc.

[0051] The image processing method provided in this application can be applied to image processing devices with data processing capabilities, such as terminal devices and servers. Terminal devices may include, but are not limited to, smartphones, desktop computers, laptops, tablets, smart speakers, in-vehicle devices, and smartwatches. Servers may be independent physical servers, server clusters or distributed systems composed of multiple physical servers, or cloud servers providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. This application does not impose specific limitations. Furthermore, the mentioned terminal devices and servers can be directly or indirectly connected via wired or wireless communication, which is also not specifically limited in this application.

[0052] The image processing devices mentioned above can possess machine learning capabilities. Machine learning (ML) is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, and many other disciplines. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to endow computers with intelligence; its applications span all areas of artificial intelligence. Machine learning and deep learning typically include techniques such as neural networks.

[0053] Figure 1 This illustrates a diagram of game screenshots provided by an existing solution. (For example...) Figure 1 As shown, in a virtual game, when the server detects suspicious behavior from an abnormal virtual player, it sends a screenshot command to the terminal device corresponding to the abnormal virtual player. Upon receiving the screenshot command from the server, the terminal device executes the command to capture the complete game interface. Then, the terminal device sends this complete game interface back to the server for processing, such as image recognition.

[0054] However, through the above Figure 1 The illustrated solution for taking screenshots in virtual games often relies on the timing of the screenshot command, which can easily lead to a mismatch between the screenshot timing and the target player's cheating behavior, preventing the terminal device from taking a screenshot at the appropriate time. For example, Figure 2 This diagram illustrates a mismatch in screenshot results when taking screenshots using existing methods. Figure 2It can be seen that when abnormal virtual players use cheating methods (such as hacks) to engage in virtual shooting in virtual games, screenshots of the game screen during this cheating should be taken (e.g., ...). Figure 2 (part (a) in the text). However, according to the screenshot command issued by the server, what was captured was the screen during the loading of a virtual game match (such as...). Figure 2 (part (b) in the text).

[0055] You can also refer to Figure 3 This illustrates a diagram of taking complete game screenshots based on existing methods. For example... Figure 3 As shown, the cheating behavior of abnormal virtual players (such as...) Figure 3 (As shown in the white dashed box in the image) It usually only occupies a small part of the complete game screen, but the traditional screenshot method captures the entire game screen of the abnormal virtual player, resulting in a lot of redundant information in the game screen captured by the traditional method, wasting a lot of transmission bandwidth and storage resources.

[0056] Therefore, in order to solve the technical problems described above, embodiments of this application provide an image processing method. This image processing method can be applied to... Figure 4 The system framework is shown. For example... Figure 4 As shown, the system framework includes a terminal device and a server. Since the code information indicates the logical code corresponding to the abnormal virtual player's behavior during the virtual game's operation, the server, after obtaining the code information, sends it to the terminal device. The terminal device can then determine the abnormal virtual player's behavior in the virtual game based on this code information, determine a screenshot strategy based on this behavior information, and then, under the guidance of the screenshot strategy, capture a pixel area of ​​a preset resolution from the virtual game interface, thereby obtaining a target pixel screenshot that reflects the abnormal gameplay situation of the abnormal virtual player.

[0057] For example, from Figure 4 It can also be seen that after the terminal device captures the target screenshot, it can also upload the pixel information corresponding to the target screenshot to the server through heartbeat messages, so that the server can realize the recognition and processing of the target screenshot.

[0058] in, Figure 4 The terminal devices shown can be understood as devices with a game client installed, such as smartphones, desktop computers, laptops, tablets, smart speakers, in-vehicle devices, smartwatches, etc., but are not specifically limited to these in this embodiment. Furthermore, the described game client can also be understood as a game application (APP).

[0059] The virtual game described can be a racing game, a shooting game, or other types of games. In practical applications, it can also include, but is not limited to, single-player games, competitive virtual games, etc. The embodiments of this application do not make any limitations.

[0060] To facilitate understanding of the technical solution of this application, an image processing method provided by an embodiment of this application will be described below with reference to the accompanying drawings.

[0061] Figure 5 A flowchart illustrating an image processing method provided in an embodiment of this application is shown. Figure 5 As shown, the image processing method may include the following steps:

[0062] 501. The server sends code information to the terminal device. The code information is used to indicate the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game.

[0063] In this example, for a specific virtual game, when an abnormal virtual player is competing against various opposing virtual players, the abnormal virtual player's behavioral state is stored in the corresponding memory of the virtual game in the form of code, thereby generating corresponding code information. The described behavioral state may include, but is not limited to, whether the abnormal virtual player is in a match, or whether they have performed a kill. This code information indicates the logical code corresponding to the abnormal virtual player's behavioral state during the virtual game's operation. Therefore, if the server cannot perceive the abnormal virtual player's behavioral state, it can obtain this code information and send it to the terminal device via dynamic code.

[0064] The described method of sending code information via dynamic code can be understood as a process where the program dynamically sends code information during the runtime of a virtual game. Sending this code information dynamically allows the overall code logic to be transmitted to the terminal device in real time, enabling the terminal device to adjust and modify the corresponding code based on the abnormal behavior of the virtual player.

[0065] It should be noted that the described terminal device can be understood as a device with a game client, such as the virtual game, installed. An abnormal virtual player can operate the virtual game through this terminal device.

[0066] 502. The terminal device determines the behavioral status information of the abnormal virtual player in the virtual game based on the code information.

[0067] In this example, after receiving the code information sent by the server, the terminal device can analyze the behavioral status information of the abnormal virtual player in the virtual game based on the code information. For instance, after receiving the code information, the terminal device can read the memory address and hook point of the logic code in the code information, thereby obtaining the behavioral status information corresponding to the memory address and hook point.

[0068] The described behavioral state information can be understood as the behavior of the abnormal virtual player in the behavioral process data stream of the virtual game that can generate game benefits. It may include a single operation behavior or a sequence of behaviors, etc., which are not limited in this application embodiment.

[0069] 503. The terminal device determines the screenshot strategy based on the behavior status information and code information.

[0070] In this example, after determining the behavior status information based on the code information, the terminal device can further determine a screenshot strategy based on the behavior status information and the code information. This screenshot strategy can be used to indicate how or at what specific time to capture a pixel area that meets a preset resolution from the game interface.

[0071] It should be noted that the described screenshot strategy may include both the screenshot time point and the screenshot location. For example, the screenshot strategy may also exclude the screenshot time point and only include the screenshot location. The following will describe in detail how to determine the screenshot strategy based on behavioral state information using different embodiments.

[0072] (a) Screenshot strategies include the time point and location of the screenshot.

[0073] For example, the terminal device can determine the screenshot time point based on behavior state information and determine the first position based on code information. The first position is the screenshot position when the target screenshot is captured from the game interface based on the screenshot time point. Then, the terminal device determines the screenshot time point and the first position as the screenshot strategy.

[0074] In this example, because an abnormal virtual player using cheating methods gains more game benefits than a normal player who doesn't use cheating, the higher the game benefits, the more likely the abnormal virtual player is cheating. Therefore, after determining the behavioral status information, the terminal device can determine the abnormal virtual player's game benefits in the virtual game based on this information. Then, the terminal device determines the time point corresponding to the maximum benefit from these benefits and uses this time point as the screenshot time point. Alternatively, in other examples, after determining the abnormal virtual player's cumulative game benefits based on the behavioral status information, the terminal device checks if the cumulative game benefits exceed a preset benefit threshold. If the cumulative game benefits exceed the preset threshold, the terminal device determines the time point corresponding to the final cumulative game benefits and uses this time point as the screenshot time point. Through these methods, a suitable screenshot timing can be identified, and a target screenshot matching the timing can be captured.

[0075] For example, Figure 6 This diagram illustrates a trigger screenshot provided in an embodiment of this application. For example... Figure 6 As shown, in a virtual game, an abnormal virtual player will generate a series of actions, forming a behavioral process data stream. Within this data stream, behavioral state information that generates benefits during the game can be detected, such as virtual firing or continuous crosshair movement. Thus, after detecting this behavioral state information, the time point corresponding to the abnormal virtual player's maximum game benefit in the virtual game can be determined, and this time point corresponding to the maximum benefit can be used as the screenshot time point.

[0076] It should be noted that the behavioral process data stream can include moving forward, backward, left, right, jumping in place, crosshair movement, virtual firing, virtual shooting, etc. In practical applications, it can also include virtual kills, driving virtual vehicles, throwing virtual bombs, etc., without further limitation. Among these, the behavioral state information includes... Figure 6 In addition to the virtual firing and continuous crosshair movement shown, practical applications may also include, but are not limited to, virtual kills, etc., which are not specifically limited in this application embodiment.

[0077] The determination of the first location can be achieved as follows: First, the terminal device can analyze the current environmental state of the abnormal virtual player in the virtual game based on the code information. For example, the terminal device can read the memory address and hook point of the environmental logic code in the code information to obtain the current environmental state corresponding to the memory address and hook point. The environmental logic code is the logic code of the environment in which the abnormal virtual player's behavior in the virtual game exists.

[0078] Then, the terminal device obtains the location coordinates of each enemy virtual player and the location coordinates of the abnormal virtual player in the virtual game based on the current environmental state. The described location coordinates can indicate the current location of the corresponding virtual player on the virtual map of the virtual game.

[0079] The terminal device further determines the suspiciousness level of each enemy virtual player based on the location coordinates of the abnormal virtual player and the location coordinates of each enemy virtual player. For example, since the probability of an enemy virtual player being outside the abnormal virtual player's field of vision is higher the further away they are, the probability of detecting the abnormal virtual player performing cheating operations against enemy virtual players at greater distances is higher in order to obtain higher game benefits. Therefore, the method for determining the suspiciousness level of each enemy virtual player can be understood as follows: After obtaining the location coordinates of the abnormal virtual player and the location coordinates of each enemy virtual player, the terminal device can calculate the coordinate difference information between the location coordinates of the abnormal virtual player and the location coordinates of each enemy virtual player, such as distance, direction, etc., which are not limited in this embodiment. In this way, the terminal device can combine the map environment information of the virtual game to prioritize each coordinate difference information, obtaining priority information. This priority information indicates the priority of each enemy virtual player's location coordinates, and then determines the suspiciousness level of each enemy virtual player based on the priority information.

[0080] For example, if enemy virtual player A is 100 meters away from an abnormal virtual player, and enemy virtual player B is 50 meters away from an abnormal virtual player, the above calculations can determine that enemy virtual player A has a suspicion level of 0.8, and enemy virtual player B has a suspicion level of 0.5. It should be understood that the 100 meters, 50 meters, etc., shown here are merely illustrative descriptions and are not intended to limit the scope of actual application.

[0081] It should be noted that the greater the coordinate difference, the higher the likelihood that the corresponding enemy virtual player is not within the abnormal virtual player's field of vision, and thus the higher the probability of detecting the abnormal virtual player engaging in cheating. Therefore, based on coordinate difference information, enemy virtual players outside the abnormal virtual player's field of vision have a higher priority than those within the field of vision. Furthermore, within the same field of vision of the abnormal virtual player, the location coordinates of enemy virtual players farther away from the abnormal virtual player have a higher priority than those closer to the abnormal virtual player.

[0082] Finally, the terminal device can determine the target suspicion level from all the suspicion levels of the opposing virtual players, and determine the location coordinates of the opposing virtual player corresponding to the target suspicion level as the first location. For example, the target suspicion level is the highest suspicion level among all the opposing virtual players. For instance, if the suspicion level of opposing virtual player A is 0.8 and the suspicion level of opposing virtual player B is 0.5, 0.8 can be taken as the target suspicion level, and the location coordinates of opposing virtual player A corresponding to 0.8 can be determined as the screenshot location, i.e., the first location.

[0083] (ii) Screenshot strategy includes screenshot location

[0084] For example, the terminal device can determine the second location based on the behavior state information and use the second location as the screenshot strategy. It should be noted that the second location is the screenshot location when capturing the target screenshot from the game interface.

[0085] In this example, besides using the time point with the maximum benefit (i.e., the screenshot time point) and the first position as the screenshot strategy, in practical applications, the screenshot strategy can also include only the screenshot position. For example, if the subsequent screenshot operation in step 504 is not performed based on the screenshot time point, it is very likely that the screenshot will be missed. In this case, the screenshot position can be directly used as the screenshot strategy to capture a pixel area that meets the preset image resolution based on the screenshot position.

[0086] In some examples, the terminal device can obtain the location coordinates of the abnormal virtual player based on the code information and determine the location coordinates of the abnormal virtual player as the second location. It should be noted that the method of determining the second location described here can be understood with reference to the process of determining the first location in (I) above, and will not be repeated here.

[0087] 504. According to the screenshot strategy, the terminal device captures a pixel area that meets the preset image resolution from the game interface of the virtual game to obtain the target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game.

[0088] In this example, after determining the screenshot strategy based on behavioral state information and code information, the terminal device can execute the screenshot operation according to the screenshot strategy to capture a pixel area that meets a preset resolution from the game interface of the virtual game, thereby obtaining the target screenshot. By capturing a pixel area that meets the preset resolution from the complete game interface, the resulting target screenshot is a small portion of the entire game interface, that is, only capturing the interface that reflects the abnormal game situation of the abnormal virtual player, thus filtering out other interfaces in the game interface besides the target pixel screenshot, such as the game background interface. It should be noted that the described abnormal game situation may include, but is not limited to, using cheating methods such as hacks to operate the game; this embodiment does not limit the description.

[0089] In some examples, when the screenshot strategy involves a screenshot time point and a first location, the terminal device can use this time point and the first location as the screenshot center to capture a pixel area from the game interface that meets a preset image resolution, thus obtaining the target screenshot. In this way, since the first location can be understood as the coordinates of the most suspicious enemy virtual player, using the first location as the screenshot center ensures that the captured target screenshot meeting the preset image resolution includes a cheating indicator, which identifies the corresponding enemy virtual player. Furthermore, the screenshot time point is determined based on the terminal device's behavior state information and code information; capturing at this time point ensures that the captured target screenshot matches the situation when the abnormal virtual player cheated.

[0090] For example, if the first position is the coordinates of the enemy virtual player A, then the center of the screenshot is the coordinates of the enemy virtual player A, and a pixel area with a preset image resolution of 100×100 is captured from the game interface to obtain the target screenshot.

[0091] It should be noted that the first location, screenshot time point, and other details described here can be understood by referring to the content in step 503 above, and will not be repeated here. Additionally, the preset image resolution determines the number of pixels contained in the captured target screenshot. For example, the preset image resolution can be set according to actual needs, such as based on the bandwidth of the data channel during data reporting. Besides the aforementioned 100×100, the preset image resolution can also be 150×150, 200×150, etc., in practical applications, and is not limited in this embodiment.

[0092] In other examples, when the screenshot strategy is set to a second location, the terminal device can use this second location as the screenshot center to capture a pixel area from the virtual game interface that meets a preset image resolution, thus obtaining the target screenshot. By using this method, since the second location is obtained from the coordinates of the abnormal virtual player, using this second location as the screenshot center ensures that the captured target screenshot largely includes enemy virtual players located near the abnormal virtual player, thereby displaying the corresponding cheating indicator in the target screenshot. Furthermore, this method avoids missing screenshots even if the screenshot operation is not performed according to the aforementioned screenshot time point.

[0093] It should be noted that the second position and other content described here can be understood with reference to the content in step 503 above, and will not be repeated here.

[0094] In this embodiment, since the code information reflects the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game, after the server sends the code information to the terminal device, the terminal device can determine the behavior state information of the abnormal virtual player in the virtual game based on the code information. Thus, the terminal device can determine a screenshot strategy based on the behavior state information and the code information, and then, under the guidance of the screenshot strategy, capture a pixel area from the game interface of the virtual game that meets a preset image resolution, thereby obtaining a target screenshot that can reflect the abnormal game situation of the abnormal virtual player in the virtual game. Through the above method, the terminal device can fully consider the behavior state information of the abnormal virtual player in the virtual game, so that the screenshot strategy determined by the behavior state information and the code information can match the cheating behavior of the abnormal virtual player. Therefore, by performing the screenshot operation using a screenshot strategy that matches the cheating behavior of the abnormal virtual player, an accurate target screenshot can be captured. Moreover, by capturing a pixel area from the game interface that meets a preset image resolution, the captured target screenshot can reflect the abnormal game situation of the abnormal virtual player in the virtual game. In other words, the target screenshot only occupies a small part of the entire game interface, that is, it only captures the cheating interface of the abnormal virtual player when abnormal game situations occur in the virtual game from the complete game interface, thereby filtering out redundant game background information and saving transmission bandwidth and storage resources.

[0095] In other examples, after the terminal device captures the target screenshot, it can not only cache the screenshot but also send it to the server as a pixel matrix for identification and processing. It should be noted that the pixel matrix described here can be understood as representing the pixel information corresponding to the target screenshot in matrix form.

[0096] For example, Figure 7 Another schematic flowchart of the image processing method provided in an embodiment of this application is shown. Figure 7 As shown, the image processing method includes at least the following steps:

[0097] 701. The server sends code information to the terminal device. The code information is used to indicate the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game.

[0098] 702. The terminal device determines the behavioral status information of the abnormal virtual player in the virtual game based on the code information.

[0099] 703. The terminal device determines the screenshot strategy based on the behavior status information and code information.

[0100] 704. According to the screenshot strategy, the terminal device captures a pixel area that meets the preset image resolution from the game interface of the virtual game to obtain the target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game.

[0101] It should be noted that the contents of steps 701 to 704 described above can be referred to in the foregoing. Figure 5 The content described in steps 501 to 504 is for your understanding and will not be repeated here.

[0102] 705. The terminal device sends the pixel information corresponding to the target screenshot to the server.

[0103] In this example, after capturing a target screenshot, the terminal device can send the pixel information corresponding to the target screenshot to the server using a pixel matrix. The described pixel information can be the RGB pixel values ​​of the target screenshot.

[0104] For example, the terminal device can map the pixel information of the target screenshot into a first message, so that the first message can contain the pixel information of the target screenshot. Then, the terminal device sends the first message to the server to achieve the purpose of sending the pixel information of the target screenshot to the server.

[0105] It should be noted that the first message described may include, but is not limited to, heartbeat messages or other types of messages, and this application embodiment does not limit the description.

[0106] 706. The server performs target image recognition processing based on pixel information.

[0107] In this example, after receiving the pixel information of the target screenshot sent by the terminal device, the server can perform identification processing on the target screenshot, such as determining whether the target screenshot is a screenshot taken when an abnormal virtual player is cheating. For example, since the terminal device maps pixel information to the first message, the server, after receiving the first message sent by the terminal device, can perform demapping processing on the first message to obtain the pixel information carried in the first message. It should be noted that the first message described here can be understood by referring to the content described in step 705 above, and will not be repeated here.

[0108] In other examples, to efficiently identify the category of the target screenshot, such as whether cheating is involved, the server can first build a target recognition model before processing the target screenshot, and then perform type identification processing based on this model. During the construction of this target recognition model, cheating characteristics in the first screenshot sample can be determined first through cheat testing or manual annotation. It should be noted that the described first screenshot sample is a screenshot taken from the game interface when the first virtual player experiences an abnormal game situation in the virtual game, meeting the preset image resolution. The specific screenshot method is described above. Figure 5 The screenshot method used for the target screenshot will be understood and will not be elaborated here. The first virtual player described can be understood as the virtual player at the time the abnormal game situation occurs, such as the abnormal virtual player described above. The described cheating characteristics may include, but are not limited to, one or more of the following characteristics: the appearance of wallhacks such as rectangular boxes, crosshair boxes, and character skeletons; the position and distribution of wallhack markers in the first screenshot sample; the size ratio of wallhack markers in the first screenshot sample; the number of wallhack markers appearing in the first screenshot sample; and the color of wallhack markers. The described wallhack markers can be used to identify enemy virtual players.

[0109] It should be noted that since the "wallhack" icon can be used to represent an enemy virtual player, its position in the first screenshot sample can be represented by the corresponding enemy virtual player's position coordinates. Furthermore, the enemy virtual player's position coordinates can be determined by a coordinate system constructed from the x and y axes. Experiments and actual cheat screenshots have statistically shown that the enemy virtual player's position coordinates (x, y) follow a normal distribution. Therefore, for details regarding the positional distribution of the aforementioned cheat features such as the "wallhack" icon in the first screenshot sample, please refer to... Figure 8A The diagram illustrating the distribution of cheating features is provided for understanding. By analyzing the positional distribution of perspective markers in the first screenshot sample, feature support can be provided for the subsequent construction of training samples.

[0110] Similarly, since the first screenshot sample is a screenshot taken from the game interface when the first virtual player experiences an abnormal game situation in the virtual game, and meets the preset image resolution, for example, each first screenshot sample is an image of size 100×100 that meets the preset image resolution. Therefore, the size of the perspective marker in the first screenshot sample is between 0 and 100. Moreover, the perspective marker of the enemy virtual player farther away from the first virtual player is smaller, and the perspective marker of the enemy virtual player closer to the first virtual player is larger. And since, as described in step 504 above, the position coordinates of the enemy virtual player who is farther away will be preferentially selected as the center of the screenshot, the size distribution of the perspective marker in the above cheating features can be specifically referred to Figure 8B The diagram illustrating the distribution of cheating features is provided for understanding. By analyzing the distribution of the size proportion of perspective markers in the first screenshot sample, feature support can be provided for the subsequent construction of training samples.

[0111] The number of enemy virtual players appearing within the field of view of the first virtual player corresponds to the number of wallhack icons in the first screenshot sample. The farther away an enemy virtual player is, the fewer enemy virtual players appear within the first virtual player's field of view, and the fewer wallhack icons appear in the first screenshot sample. Therefore, the number of wallhack icons in the first screenshot sample decreases exponentially with increasing distance, as detailed in [reference needed]. Figure 8C The diagram illustrating the distribution of cheating features is provided for understanding. By analyzing the distribution of the number of occurrences of perspective markers in the first screenshot sample, feature support can be provided for the subsequent construction of training samples.

[0112] Furthermore, the colors of the perspective markers in the cheating features described above follow a random distribution; for example, the colors of the perspective markers could be red, yellow, blue, green, etc. Analyzing the color distribution of the perspective markers can provide feature support for the subsequent construction of training samples.

[0113] For example, the server can select training samples based on various different methods during the construction of training samples. For instance, Figure 9 A schematic diagram illustrating the construction method of the training samples provided in the embodiments of this application is shown. For example... Figure 9As shown, all screenshot samples from the "white account" are obtained, and feature analysis is performed on all screenshot samples. Then, screenshot samples containing cheating features, such as those mentioned above, are filtered out, and random samples are drawn from those that do not contain cheating features to obtain "white samples." It should be noted that the "white samples" mentioned here can be understood as screenshots taken by the second virtual player from the game interface that meet the preset image resolution when no abnormal game situation occurs in the virtual game; i.e., the second screenshot samples. The described second virtual player can be understood as a normal player who has not performed cheating operations, such as an enemy virtual player who has not performed cheating operations.

[0114] Similarly, black samples can be obtained in the following three ways: ① Filtering screenshots by testing and penalizing accounts, and selecting screenshots of virtual players found to be cheating through manual screening; ② Screening cheating screenshots with solid or dashed boxes by automatically constructing a large number of bounding boxes; ③ Screening screenshots containing features such as wallhacks by combining GAN anomaly detection with manual screening. It should be noted that the described black samples can be understood as the first type of screenshot sample described above.

[0115] For example, using method ② above to filter out cheating screenshots with solid or dashed boxes, Figure 10 This illustration shows a schematic diagram of constructing a black sample according to an embodiment of this application. Figure 10 As shown, since the target screenshot sample is a screenshot taken from the game interface by the first virtual player when the abnormal game situation occurs in the virtual game, and it meets the preset image resolution, after the server analyzes and determines the corresponding cheating features from the target screenshot sample, it constructs the first screenshot sample by superimposing these cheating features onto the second screenshot sample. For example, by superimposing these cheating features onto the second screenshot sample, corresponding tag data can also be constructed, which can be used to distinguish between black samples and white samples.

[0116] For example, let's take the screenshots that contain perspective cheats as an example, filtered out using method ③ above, to illustrate how to obtain the corresponding first screenshot sample. Figure 11 This diagram illustrates the process of selecting the first screenshot sample based on a GAN model. Because the GAN model learns the feature distribution of the second screenshot sample (i.e., the normal sample) through its generator and discriminator, if the feature distribution of certain screenshot samples, after being determined by the GAN model, differs from that of the second screenshot sample, it indicates that the corresponding screenshot sample is an anomaly, thus allowing the selection of the first screenshot sample. For example, such as... Figure 11As shown in part (a) of the diagram, during the prediction phase, since the features of the second screenshot sample have been learned, encoder 1 encodes the second screenshot sample X to obtain a feature vector z. The decoder then decodes the feature vector z to generate the original image, i.e., sample X'. Encoder 2 then encodes this sample X' to generate the corresponding feature vector z'. By comparing the differences between feature vector z' and feature vector z, it can be determined that feature vector z' and feature vector z are similar, indicating that sample X' is not an anomalous sample.

[0117] Based on the same idea, since the GAN model did not learn the features of the first screenshot sample (anomaly sample), encoder 1 encodes the first screenshot sample X1 to obtain the feature vector Z1. The decoder then decodes the feature vector Z1 to generate the original image, i.e., sample X1'. Encoder 2 then encodes this sample X1' to generate the corresponding feature vector Z1'. By comparing the difference between feature vector Z1' and feature vector Z1, it can be determined that if there is a significant difference between them, sample X1' is an anomaly sample. Therefore, in this GAN model, by setting a threshold, anomaly samples are filtered out from a large number of screenshot samples, thus identifying the first screenshot sample.

[0118] from Figure 11 As shown in section (b), the results show that the original screenshot of the second screenshot sample (i.e., the normal sample) is identical to the generated screenshot, with negligible differences. However, the screenshot of the first screenshot sample (i.e., the abnormal sample) and the corresponding generated screenshot both show signs of cheating, such as traces of perspective manipulation in the boxes and figures.

[0119] Finally, we can further identify all the first screenshot samples through simple manual filtering, and find some screenshot samples from the second screenshot samples that resemble cheating features. By using GAN anomaly detection combined with manual screening to filter screenshot samples, we can effectively filter out abnormal samples from a large range of screenshot samples, and we can also filter out screenshot samples with various cheating features, such as wallhacks, solid or dashed box cheats, character skeletons, etc.

[0120] In this way, the server can construct training samples based on the first and second screenshot samples.

[0121] For example, after constructing the training samples, if some misclassified screenshot samples are found, such as screenshots with similar boxes or bright colors, then the training samples need to be expanded and refined. For instance, Figure 12A flowchart illustrating the sample expansion and purification process provided in this application embodiment is shown. The server can train multiple classification models with different parameters using training samples, such as the Resnetxt model and the Mobilenet-v3 large model. By training these multiple classification models with different parameters, it is possible to filter out anomalous samples from a large number of screenshot samples that match the model parameters of each classification model. Furthermore, combining these multiple classification models with different parameters improves the accuracy of anomalous sample filtering. Therefore, the server predicts test data based on the trained multiple classification models. If the predicted labels of the test data are the same, they are automatically labeled; otherwise, if the predicted labels of the test data are different, further manual correction is performed to obtain more anomalous samples, i.e., more first screenshot samples. In this way, the server adds the anomalous samples obtained through expansion and purification to the training samples, thus achieving sample expansion of the training samples.

[0122] In addition, after Figure 12 Even after sample augmentation and purification as shown in the diagram, noisy samples may still exist in the training samples. In this case, for example... Figure 13 The purification process is performed using the five-fold cross-validation method shown in the figure. Figure 13 A schematic diagram of the sample purification process provided in an embodiment of this application is shown. For example... Figure 13 As shown, taking the five-fold cross-validation method as an example, after... Figure 12 After processing, the abnormal samples are used to train the model by repeatedly selecting different four-fold samples from the five-fold sample set and then predicting the remaining one-fold sample set. This process is repeated five times before a prediction is made on the entire sample set. In this way, the server can formulate a strategy based on the final predicted sample situation to correct the labels of abnormal samples, thereby ensuring the purity of the training samples.

[0123] In this way, after obtaining the training samples, the server can train the initial recognition model based on the training samples. For example, the server can iteratively train the initial recognition model based on the first screenshot sample and the second screenshot sample to obtain the target recognition model. It should be noted that the described initial recognition model may include, but is not limited to, the MobileNet-V3 Large model, other binary classification models, etc., and is not limited to these in this embodiment.

[0124] For example, the Mobilenet-v3 large model can be understood as a lightweight binary classification model based on grouped classification convolutions of pixel screenshots. Figure 14 A schematic diagram of the structure of the Mobilenet-v3 large model provided in an embodiment of this application is shown. Figure 14 As shown, the grouping and separating convolution parameters in the Mobilenet-v3 large model are 88. Specifically, the server can input the sample features of the first screenshot sample and the sample features of the second screenshot sample into the convolutional layer of the Mobilenet-v3 large model. The convolutional layer performs 3×3 grouping convolution processing to achieve partial fusion of the screenshot samples. Then, the sample features of the partially fused screenshot samples are subjected to 1×1 convolution processing to achieve global fusion of the screenshot samples, thereby training the target recognition model.

[0125] In this way, after training the target recognition model, the server can use this model to recognize and process the pixel information of the target screenshot, thereby identifying the category of the target screenshot. For example, if the server determines that the target screenshot indicates an abnormal game situation, it can then impose game penalties or other actions on the abnormal virtual player.

[0126] It should be noted that after the target recognition model is trained on the server, in addition to performing screenshot recognition processing on the server side, the screenshot recognition processing described in step 706 can also be deployed and executed on the terminal device side in practical applications. For example, after the server trains the target recognition model, it can distribute the target recognition model to the terminal device, and then, after the terminal device captures a screenshot of the target, it can perform target recognition processing on the screenshot based on the target recognition model.

[0127] By analyzing the existing cheating features in the first screenshot sample and artificially simulating cheating effects in the second screenshot sample, a target recognition model is trained. This model can then efficiently identify the category of the target screenshot. Furthermore, by constructing corresponding label data from the first and second screenshot samples, the cold start problem caused by insufficient label data can be solved.

[0128] In other examples, the server can use the identification of cheating and normal actions by virtual players in the virtual game as training objectives, and use abnormal samples of virtual players in the virtual game as training samples to iteratively train and obtain a first recognition model. After training the first recognition model, the server can also send it to the terminal device via dynamic code. Upon receiving the first recognition model, the terminal device can identify the abnormal actions of the virtual player in the virtual game, and when the action is identified as cheating, directly extract a pixel area from the game interface that meets a preset image resolution to obtain the target screenshot. This method improves the overall coverage of cheating behavior. It should be noted that the cheating behaviors described here include, but are not limited to, virtual firing and continuous crosshair movement.

[0129] In other examples, when a normal player takes a screenshot using the aforementioned screenshot strategy, the captured image may only be a background image of the virtual game environment, excluding screenshots of abnormal game conditions. Therefore, after capturing image A using the aforementioned screenshot strategy, the normal player's terminal device can wait for a preset time t and then capture image B from the same game perspective. The difference between each pixel in image A and each pixel at the same location in image B is calculated, as detailed in [reference needed]. Figure 15 The diagram shown is for reference. If the sum of the differences is less than a preset threshold, then image A is considered to be the same as image B. In this case, when the terminal device reports the first message, it may not report images A and B, but only report a screenshot without suspicion, to indicate that there is no abnormal game situation in the screenshot.

[0130] The foregoing primarily describes the solutions provided by the embodiments of this application from a methodological perspective. It is understood that to achieve the above functions, corresponding hardware structures and / or software modules are included to execute each function. Those skilled in the art should readily recognize that, based on the modules and algorithm steps described in conjunction with the embodiments disclosed in this application, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0131] This application embodiment can divide the device into functional modules according to the above method example. For example, each function can be divided into its own functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0132] The terminal device in the embodiments of this application will be described in detail below. Figure 16 This is a schematic diagram of one embodiment of the terminal device provided in this application. Figure 16 As shown, the terminal device may include an acquisition unit 1601 and a processing unit 1602. Exemplarily, the terminal device may also include a sending unit 1603.

[0133] The acquisition unit 1601 is used to acquire code information sent by the server. This code information indicates the logic code corresponding to the behavioral state of the abnormal virtual player during the operation of the virtual game. The processing unit 1602 is used to: determine the behavioral state information of the abnormal virtual player in the virtual game based on the code information; determine a screenshot strategy based on the behavioral state information and the code information; and, according to the screenshot strategy, extract a pixel area from the game interface of the virtual game that meets a preset image resolution to obtain a target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game.

[0134] In some optional implementations, the processing unit 1602 is configured to: determine the screenshot time point based on the behavior state information, and determine the first position based on the code information, wherein the first position is the screenshot position when the target screenshot is captured from the game interface based on the screenshot time point; determine the screenshot time point and the first position as a screenshot strategy; and, based on the screenshot time point and with the first position as the screenshot center, capture a pixel area that meets the preset image resolution from the game interface of the virtual game to obtain the target screenshot.

[0135] In some alternative implementations, the processing unit 1602 is configured to: determine the game gains of the abnormal virtual player in the virtual game based on the behavior status information; determine the time point corresponding to the maximum gain in the game gains; and determine the time point corresponding to the maximum gain as the screenshot time point.

[0136] In some alternative embodiments, the acquisition unit 1601 is configured to: acquire the position coordinates of each enemy virtual player in the virtual game, and the position coordinates of abnormal virtual players, based on code information. The processing unit 1602 is configured to: determine the suspiciousness level of each enemy virtual player based on the position coordinates of the abnormal virtual players and the position coordinates of each enemy virtual player; determine the target suspiciousness level from the suspiciousness levels of each enemy virtual player; and the terminal device determines the position coordinates of the enemy virtual player corresponding to the target suspiciousness level as the first position.

[0137] In some alternative embodiments, the processing unit 1602 is configured to: calculate the coordinate difference information between the location coordinates of the abnormal virtual player and the location coordinates of each enemy virtual player; prioritize each coordinate difference information to obtain priority information, which is used to indicate the priority of the location coordinates of each enemy virtual player; and determine the degree of suspicion of each enemy virtual player based on the priority information.

[0138] In some alternative embodiments, the processing unit 1602 is configured to: determine a second position based on code information, the second position being the screenshot position when capturing a target screenshot from the game interface; determine the second position as a screenshot strategy; and, with the second position as the screenshot center, capture a pixel area from the game interface of the virtual game that meets a preset image resolution to obtain a target screenshot.

[0139] In some alternative embodiments, the acquisition unit 1601 is used to: acquire the location coordinates of the abnormal virtual player based on code information. The processing unit 1602 is used to determine the location coordinates of the abnormal virtual player as a second location.

[0140] In some alternative embodiments, the sending unit 1603 is configured to: after capturing a pixel area that meets a preset image resolution from the game interface of the virtual game according to the screenshot strategy and obtaining a target screenshot, send the pixel information corresponding to the target screenshot to the server, and the pixel information is used by the server to identify and process the target screenshot.

[0141] In some alternative implementations, the sending unit 1603 is configured to: send a first message to the server, the first message carrying pixel information corresponding to the target screenshot.

[0142] The server in the embodiments of this application will be described in detail below. Figure 17 This is a schematic diagram of one embodiment of the server provided in this application. Figure 17 As shown, the server may include a sending unit 1701. Exemplarily, the server may also include a processing unit 1702 and an acquisition unit 1703.

[0143] The sending unit 1701 is used to send code information to the terminal device. The code information is used to indicate the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game. The code information is used to instruct the terminal device to determine the behavior state information of the abnormal virtual player in the virtual game. The behavior state information is used to instruct the terminal device to determine a screenshot strategy. The screenshot strategy is used to instruct the terminal device to capture a pixel area that meets a preset image resolution from the game interface of the virtual game to obtain a target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game.

[0144] In some embodiments, the acquisition unit 1703 is further configured to receive pixel information corresponding to the target screenshot sent by the terminal device after sending code information to the terminal device. The processing unit 1702 is configured to perform identification processing on the target screenshot based on the pixel information.

[0145] In some alternative embodiments, the acquisition unit 1703 is further configured to acquire a first screenshot sample and a second screenshot sample before performing recognition processing on the target screenshot based on the pixel information. The first screenshot sample is a screenshot taken from the game interface by a first virtual player when the abnormal game situation occurs in the virtual game, satisfying the preset image resolution. The second screenshot sample is a screenshot taken from the game interface by a second virtual player when the abnormal game situation does not occur in the virtual game, satisfying the preset image resolution. The processing unit 1702 is further configured to: iteratively train an initial recognition model based on the first screenshot sample and the second screenshot sample to obtain a target recognition model; and process the pixel information based on the target recognition model to perform recognition processing on the target screenshot to obtain the category of the target screenshot.

[0146] In some alternative embodiments, the processing unit 1702 is configured to: determine cheating features in a target screenshot sample, wherein the target screenshot sample is a screenshot taken from the game interface by the first virtual player when the abnormal game situation occurs in the virtual game, and the screenshot meets the preset image resolution; and superimpose the cheating features onto the second screenshot sample to obtain the first screenshot sample.

[0147] In some alternative embodiments, the acquisition unit 1703 is used to receive a first message sent by the terminal device, the first message carrying pixel information corresponding to the target screenshot.

[0148] The image processing apparatus in the embodiments of this application has been described above from the perspective of modular functional entities. The image processing apparatus in the embodiments of this application will now be described below from the perspective of hardware processing. The described image processing apparatus may be... Figure 16 The terminal device shown, or Figure 17 The servers shown are examples of this. Figure 18 This is a schematic diagram of the structure of an image processing apparatus provided in an embodiment of this application. The image processing apparatus can vary considerably depending on its configuration or performance. The image processing apparatus may include at least one processor 1801, a communication line 1807, a memory 1803, and at least one communication interface 1804.

[0149] The processor 1801 may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (server IC), or one or more integrated circuits used to control the execution of programs according to the present application.

[0150] Communication line 1807 may include a path for transmitting information between the aforementioned components.

[0151] The communication interface 1804 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0152] The memory 1803 can be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions. The memory can exist independently and be connected to the processor via communication line 1807. The memory can also be integrated with the processor.

[0153] The memory 1803 stores computer execution instructions for implementing the scheme of this application, and its execution is controlled by the processor 1801. The processor 1801 executes the computer execution instructions stored in the memory 1803, thereby implementing the method provided in the above embodiments of this application.

[0154] Optionally, the computer execution instructions in the embodiments of this application may also be referred to as application code, and the embodiments of this application do not specifically limit this.

[0155] In a specific implementation, as one example, the image processing device may include multiple processors, such as... Figure 18 Processors 1801 and 1802 are mentioned. Each of these processors can be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. A processor here can refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0156] In a specific implementation, as one embodiment, the image processing apparatus may further include an output device 1805 and an input device 1806. The output device 1805 communicates with the processor 1801 and can display information in various ways. The input device 1806 communicates with the processor 1801 and can receive input from a target object in various ways. For example, the input device 1806 may be a mouse, a touchscreen device, or a sensing device, etc.

[0157] The image processing device described above can be a general-purpose device or a special-purpose device. In specific implementations, the image processing device can be a server, terminal equipment, or other similar devices. Figure 18 A device with a similar structure. The embodiments of this application do not limit the type of image processing device.

[0158] It should be noted that Figure 18 The processor 1801 can invoke computer execution instructions stored in memory 1803 to cause the image processing device to perform actions such as... Figure 5 or Figure 7 The method in the corresponding method embodiment.

[0159] Specifically, Figure 16 Processing unit 1602 in Figure 17 The function / implementation process of the processing unit 1702 can be achieved through... Figure 18 The processor 1801 in the memory calls computer execution instructions stored in memory 1803 to achieve this. Figure 16 The acquisition unit 1601 and the sending unit 1603 in the middle Figure 17 The functions / implementation process of the sending unit 1701 and the acquiring unit 1703 can be obtained through Figure 18 It is implemented using the 1804 communication interface.

[0160] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the image processing methods described in the above method embodiments.

[0161] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the image processing methods described in the above method embodiments.

[0162] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product.

[0163] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0164] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0165] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0166] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0167] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0168] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, they can be implemented in whole or in part in the form of a computer program product.

[0169] A computer program product includes one or more computer instructions. When these computer instructions are loaded and executed on a computer, they generate, in whole or in part, the processes or functions according to embodiments of this application. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., SSDs), etc.

[0170] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An image processing method, characterized in that, include: The terminal device obtains code information sent by the server, which is used to indicate the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game; The terminal device determines the behavioral status information of the abnormal virtual player in the virtual game based on the code information; The terminal device determines the screenshot strategy based on the behavior status information and the code information; According to the screenshot strategy, the terminal device captures a pixel area that meets the preset image resolution from the game interface of the virtual game to obtain a target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game, and the target screenshot is smaller than the game interface.

2. The method according to claim 1, characterized in that, The terminal device determines a screenshot strategy based on the behavior status information and the code information, including: The terminal device determines the screenshot time point based on the behavior state information and determines the first position based on the code information. The first position is the screenshot position when the target screenshot is captured from the game interface based on the screenshot time point. The terminal device determines the screenshot time point and the first location as the screenshot strategy; The terminal device, according to the screenshot strategy, extracts a pixel region that meets a preset image resolution from the game interface of the virtual game to obtain a target screenshot, including: The terminal device, based on the screenshot time point, takes the first position as the screenshot center and extracts a pixel area from the game interface of the virtual game that meets the preset image resolution to obtain the target screenshot.

3. The method according to claim 2, characterized in that, The terminal device determines the screenshot time point based on the behavior status information, including: The terminal device determines the game rewards of the abnormal virtual player in the virtual game based on the behavioral status information. The terminal device determines the time point corresponding to the maximum gain in the game rewards; The terminal device determines the time point corresponding to the maximum benefit as the screenshot time point.

4. The method according to claim 2 or 3, characterized in that, The terminal device determines the first location based on the code information, including: The terminal device obtains the location coordinates of each enemy virtual player in the virtual game, as well as the location coordinates of the abnormal virtual player, based on the code information. The terminal device determines the suspiciousness level of each enemy virtual player based on the location coordinates of the abnormal virtual player and the location coordinates of each enemy virtual player; The terminal device determines the target's suspiciousness level from the suspiciousness level of each of the enemy virtual players; The terminal device determines the location coordinates of the enemy virtual player corresponding to the suspiciousness level of the target as the first location.

5. The method according to claim 4, characterized in that, The terminal device determines the suspiciousness level of each enemy virtual player based on the location coordinates of the abnormal virtual player and the location coordinates of each enemy virtual player, including: The terminal device calculates the coordinate difference information between the position coordinates of the abnormal virtual player and the position coordinates of each enemy virtual player; The terminal device sorts each coordinate difference information by priority to obtain priority information, which is used to indicate the priority of the position coordinates of each enemy virtual player. The terminal device determines the degree of suspicion of each enemy virtual player based on the priority information.

6. The method according to claim 1, characterized in that, The terminal device determines a screenshot strategy based on the behavior status information and code information, including: The terminal device determines a second position based on the code information, the second position being the screenshot position when the target screenshot is captured from the game interface; The terminal device determines the second location as the screenshot strategy; The terminal device, according to the screenshot strategy, extracts a pixel region that meets a preset image resolution from the game interface of the virtual game to obtain a target screenshot, including: The terminal device uses the second position as the center of the screenshot and extracts a pixel area that meets the preset image resolution from the game interface of the virtual game to obtain the target screenshot.

7. The method according to claim 6, characterized in that, The terminal device determines the second location based on the code information, including: The terminal device obtains the location coordinates of the abnormal virtual player based on the code information; The terminal device determines the location coordinates of the abnormal virtual player as the second location.

8. The method according to any one of claims 1 to 3, 6 to 7, characterized in that, After the terminal device extracts a pixel region meeting a preset image resolution from the game interface of the virtual game according to the screenshot strategy to obtain the target screenshot, the method further includes: The terminal device sends pixel information corresponding to the target screenshot to the server, and the pixel information is used by the server to identify and process the target screenshot.

9. The method according to claim 8, characterized in that, The terminal device sends the pixel information corresponding to the target screenshot to the server, including: The terminal device sends a first message to the server, the first message carrying pixel information corresponding to the target screenshot.

10. An image processing method, characterized in that, include: The server sends code information to the terminal device. The code information is used to indicate the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game. The code information is used to instruct the terminal device to determine the behavior state information of the abnormal virtual player in the virtual game. The behavior state information is used to instruct the terminal device to determine a screenshot strategy. The screenshot strategy is used to instruct the terminal device to capture a pixel area that meets a preset image resolution from the game interface of the virtual game to obtain a target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game. The target screenshot is smaller than the game interface.

11. The method according to claim 10, characterized in that, After the server sends the code information to the terminal device, the method further includes: The server receives pixel information corresponding to the target screenshot sent by the terminal device; The server performs identification processing on the target screenshot based on the pixel information.

12. The method according to claim 11, characterized in that, Before the server performs recognition processing on the target screenshot based on the pixel information, the method further includes: The server acquires a first screenshot sample and a second screenshot sample, wherein the first screenshot sample is a screenshot taken from the game interface by the first virtual player when the abnormal game situation occurs in the virtual game, and the second screenshot sample is a screenshot taken from the game interface by the second virtual player when the abnormal game situation does not occur in the virtual game, and the second screenshot sample is a screenshot taken from the game interface by the second virtual player, and the second screenshot sample is a screenshot taken from the game interface by the second virtual player when the abnormal game situation does not occur, and the second screenshot sample is a screenshot taken from the game interface, and the second screenshot sample is a screenshot taken from the game interface, and the second screenshot sample is a screenshot taken from the game interface by the second virtual player, and the second screenshot sample is a screenshot taken from the game interface ... The server iteratively trains the initial recognition model based on the first screenshot sample and the second screenshot sample to obtain the target recognition model; The server performs identification processing on the target screenshot based on the pixel information, including: The server processes the pixel information based on the target recognition model to identify the target screenshot and obtain the category of the target screenshot.

13. The method according to claim 12, characterized in that, The server obtains the first screenshot sample, including: The server determines the cheating features in the target screenshot sample, which is a screenshot taken from the game interface by the first virtual player when the abnormal game situation occurs in the virtual game, and which meets the preset image resolution. The server overlays the cheating features onto the second screenshot sample to obtain the first screenshot sample.

14. The method according to any one of claims 11 to 13, characterized in that, The server receives pixel information corresponding to the target screenshot sent by the terminal device, including: The server receives a first message sent by the terminal device, the first message carrying pixel information corresponding to the target screenshot.

15. A terminal device, characterized in that, include: The acquisition unit is used to acquire code information sent by the server, the code information being used to indicate the logic code corresponding to the behavior state of the abnormal virtual player during the operation of the virtual game; The processing unit is used to determine the behavioral status information of the abnormal virtual player in the virtual game based on the code information; The processing unit is used to determine a screenshot strategy based on the behavior status information and the code information; The processing unit is configured to extract a pixel region that meets a preset image resolution from the game interface of the virtual game according to the screenshot strategy, and obtain a target screenshot. The target screenshot is used to reflect the abnormal game situation of the abnormal virtual player in the virtual game, and the target screenshot is smaller than the game interface.

16. An image processing apparatus, characterized in that, The image processing device includes: an input / output (I / O) interface, a processor, and a memory, wherein the memory stores program instructions; The processor is configured to execute program instructions stored in the memory, perform the method as described in any one of claims 1 to 9, or perform the method as described in any one of claims 10 to 14.

17. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes instructions that, when executed on a computer device, cause the computer device to perform the method as claimed in any one of claims 1 to 9, or to perform the method as claimed in any one of claims 10 to 14.

18. A computer program product, characterized in that, The finished computer program includes instructions that, when executed on a computer device or processor, cause the computer device or processor to perform the method as described in any one of claims 1 to 9, or to perform the method as described in any one of claims 10 to 14.

Images