A method, device, computer device and storage medium for detecting game cheats

Abstract

An embodiment of the present application discloses a game cheating detection method, device, computer device, and storage medium. In this solution, multiple sample cheating interface images are obtained; based on the co-occurrence relationship among the text contents of the sample cheating interface images, multiple associated phrases are extracted from each sample cheating interface image; according to the frequency of each associated phrase appearing in the multiple sample cheating interface images, the cheating keyword score of each associated phrase is determined; based on the cheating keyword scores, cheating keyword groups are determined from all the associated phrases; and based on the cheating keyword groups, cheating detection is performed on candidate cheating interface images in the game interface to be detected. In this way, the accuracy of game cheating detection can be improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, and particularly to a method and device for detecting game cheats, a computer device, and a storage medium. Background Art

[0002] With the continuous popularization of the Internet and the booming development of online games, online games have gradually become a part of people's entertainment life. However, there are problems with game cheats in games. Game cheats refer to cheating programs that use computer technology to specifically target one or more online games and modify some programs in the game. Game cheats seriously damage the authenticity, accuracy, and fairness of game data.

[0003] In related technologies, game cheats are identified by taking screenshots of the game interface and using a method based on deep learning object detection for cheat identification and classification. Specifically, the object detection method labels various targets based on the features of various game cheats (or based on different game cheats), and conducts targeted learning and detection on the features of each target. However, different game scenarios require different samples to be labeled to implement a targeted model, and new game cheats need to be manually labeled and the cheat samples need to be updated and maintained, which affects the efficiency of game cheat detection. Summary of the Invention

[0004] Embodiments of this application provide a method and device for detecting game cheats, a computer device, and a storage medium, which can improve the accuracy of game cheat detection.

[0005] Embodiments of this application provide a method for detecting game cheats, including:

[0006] Obtaining a plurality of sample cheat interface images;

[0007] Extracting a plurality of associated word groups from each sample cheat interface image based on the accompanying relationship in the text content of the sample cheat interface image;

[0008] Determining the cheat keyword score of each associated word group according to the frequency of occurrence of each associated word group in the plurality of sample cheat interface images;

[0009] Determining a cheat keyword group from all the associated word groups based on the cheat keyword scores;

[0010] Performing game cheat detection on candidate cheat interface images in the game interface to be detected based on the cheat keyword group.

[0011] Correspondingly, embodiments of this application also provide a device for detecting game cheats, including:

[0012] An obtaining unit, configured to obtain a plurality of sample cheat interface images;

[0013] An extraction unit, configured to extract a plurality of associated phrase groups from each sample external plug-in interface image based on the accompanying relationship in the text content of the sample external plug-in interface image;

[0014] A first determination unit, configured to determine an external plug-in keyword score for each associated phrase group according to the frequency of occurrence of each associated phrase group in the plurality of sample external plug-in interface images;

[0015] A second determination unit, configured to determine an external plug-in keyword group from all the associated phrase groups based on the external plug-in keyword scores;

[0016] A detection unit, configured to perform an external plug-in detection on a candidate external plug-in interface image in the game interface to be detected based on the external plug-in keyword group.

[0017] In some embodiments, the first determination unit includes:

[0018] A first determination subunit, configured to determine an initial external plug-in keyword score corresponding to each associated phrase group according to the frequency of occurrence of each associated phrase group in the plurality of sample external plug-in interface images;

[0019] A second determination subunit, configured to determine an external plug-in keyword score for each associated phrase group according to the universality of each associated phrase group in non-external plug-in interface images and the initial external plug-in keyword score.

[0020] In some embodiments, the second determination subunit is specifically configured to:

[0021] Obtain a non-external plug-in keyword group set corresponding to the non-external plug-in interface images;

[0022] Determine an external plug-in keyword score suppression coefficient for each associated phrase group based on the universality of each associated phrase group in the non-external plug-in keyword group set;

[0023] Calculate a product value of the initial external plug-in keyword score and the external plug-in keyword score suppression coefficient to obtain the external plug-in keyword score of the associated phrase group.

[0024] In some embodiments, the extraction unit includes:

[0025] An identification unit, configured to perform text recognition on each sample external plug-in interface image to obtain the text corresponding to each sample external plug-in interface image;

[0026] A third determination subunit, configured to determine a word combination composed of at least two words that appear concomitantly from each text;

[0027] A fourth determination subunit, configured to obtain the plurality of associated phrase groups in each text based on the word combinations in each text.

[0028] In some embodiments, the second determination unit includes:

[0029] A fifth determination subunit, configured to screen out associated phrases with an external plug-in keyword score greater than a preset score from all the associated phrases extracted from each sample external plug-in interface image, so as to obtain the external plug-in keyword group.

[0030] In some embodiments, the detection unit includes:

[0031] An acquisition subunit, configured to acquire a game interface to be detected, and identify a non-game interface image from the game interface to be detected, so as to obtain the candidate external plug-in interface image;

[0032] A first extraction subunit, configured to extract a plurality of candidate associated phrases from the candidate external plug-in interface image based on an accompanying relationship in the text content of the candidate external plug-in interface image;

[0033] A matching subunit, configured to match the plurality of candidate associated phrases with the external plug-in keyword group;

[0034] A sixth determination subunit, configured to determine an external plug-in detection result of the game interface to be detected based on a matching result.

[0035] In some embodiments, the sixth determination subunit is specifically configured to:

[0036] If there is a candidate associated phrase that matches the external plug-in keyword group among the plurality of candidate associated phrases, determine that the external plug-in detection result of the game interface to be detected is using an external plug-in;

[0037] If there is no candidate associated phrase that matches the external plug-in keyword group among the plurality of candidate associated phrases, determine that the external plug-in detection result of the game interface to be detected is not using an external plug-in.

[0038] In some embodiments, the apparatus further includes:

[0039] An adding unit, configured to add the candidate external plug-in interface image in the game interface to be detected to a set of potential external plug-in interface images if the external plug-in detection result of the game interface to be detected indicates not using an external plug-in;

[0040] A third determination unit, configured to determine a new external plug-in keyword group according to the potential external plug-in interface images in the set of potential external plug-in interface images;

[0041] An updating unit, configured to update the external plug-in keyword group according to the new external plug-in keyword group, and perform external plug-in detection on the candidate external plug-in interface image in the game interface to be detected based on the updated external plug-in keyword group.

[0042] In some embodiments, the third determination unit includes:

[0043] A processing subunit, configured to perform clustering processing on all potential external plug-in interface images in the set of potential external plug-in interface images, to obtain a plurality of image subsets, where each image subset includes potential external plug-in interface images belonging to the same category;

[0044] An elimination subunit, configured to eliminate image subsets that are not external plug-in interface images from the plurality of image subsets, to obtain a target image subset;

[0045] A second extraction subunit, configured to extract a plurality of potential associated phrases from each potential external plug-in interface image in the target image subset;

[0046] A seventh determination subunit, configured to determine a potential external plug-in keyword score for each potential associated phrase according to the frequency of occurrence of each potential associated phrase in the potential external plug-in interface images in the target image subset;

[0047] An eighth determination subunit, configured to determine the new external plug-in keyword group from the plurality of potential associated phrases based on the potential external plug-in keyword score.

[0048] Correspondingly, an embodiment of the present application further provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the game external plug-in detection method provided in any embodiment of the present application.

[0049] Correspondingly, an embodiment of the present application further provides a storage medium, storing a plurality of instructions, where the instructions are suitable for being loaded by a processor to execute the game external plug-in detection method as described above.

[0050] In the embodiment of the present application, by obtaining a plurality of sample external plug-in interface images; extracting a plurality of associated phrases from each sample external plug-in interface image based on the accompanying relationship in the text content of the sample external plug-in interface images; determining an external plug-in keyword score for each associated phrase according to the frequency of occurrence of each associated phrase in the plurality of sample external plug-in interface images; determining an external plug-in keyword group from all the associated phrases based on the external plug-in keyword score; and performing external plug-in detection on candidate external plug-in interface images in the game interface to be detected based on the external plug-in keyword group, thereby improving the accuracy of game external plug-in detection. Description of the Drawings

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0052] Figure 1A schematic flowchart of a game cheating detection method provided by an embodiment of the present application.

[0053] Figure 2 A schematic flowchart of another game cheating detection method provided by an embodiment of the present application.

[0054] Figure 3 A structural block diagram of a game cheating detection device provided by an embodiment of the present application.

[0055] Figure 4 A schematic structural diagram of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0056] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present application.

[0057] The embodiments of the present application provide a game cheating detection method, device, storage medium, and computer device. Specifically, the game cheating detection method in the embodiments of the present application can be executed by a computer device, where the computer device can be a terminal or a server, etc. The terminal can be a smart phone, a tablet computer, a notebook computer, a touch screen, a personal computer (PC), a personal digital assistant (PDA), and other terminal devices. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

[0058] For example, the computer device can be a server, and the server can obtain multiple sample cheating interface images; extract multiple associated phrases from each sample cheating interface image based on the accompanying relationship in the text content of the sample cheating interface images; determine the cheating keyword score of each associated phrase according to the frequency of each associated phrase appearing in multiple sample cheating interface images; determine a cheating keyword group from all the associated phrases based on the cheating keyword scores; and perform cheating detection on candidate cheating interface images in the game interface to be detected based on the cheating keyword group.

[0059] Based on the above problems, the embodiments of the present application provide a first game cheating detection method, device, computer device, and storage medium, which can improve the accuracy of game cheating detection.

[0060] The following will be described in detail respectively. It should be noted that the description order of the following embodiments does not limit the preferred order of the embodiments.

[0061] The embodiments of the present application provide a game cheating detection method, which can be executed by a terminal or a server. The embodiments of the present application will be described by taking the game cheating detection method executed by the server as an example.

[0062] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of a game cheating detection method provided by the embodiments of the present application. The specific process of this game cheating detection method can be as follows:

[0063] 101. Obtain multiple sample cheating interface images.

[0064] In the embodiments of the present application, the cheating interface refers to the interface that appears on the game interface of the player's terminal using the game cheating, which is used to assist the player in game operations. The cheating interface can include multiple game functions, and different game functions can be used to assist the player in strengthening game operations in the game.

[0065] Specifically, a cheating program generally refers to a program that, during the operation of the terminal, is hooked into the space of another program through a certain event trigger (common trigger events include keyboard trigger, mouse trigger, message trigger, etc.). The purpose of hooking is usually to change the running mode of the hooked program. A game cheating program is to hook the cheating program into the game program and enhance various functions by intercepting and modifying the data sent from the game to the game server.

[0066] Among them, the sample cheating interface image refers to the image of a sample cheating interface obtained from the sample cheating interface set, and the sample cheating interface set includes multiple sample cheating interfaces. There are various ways to obtain the sample cheating interface. For example, the sample cheating interface can be the cheating interface that appears on the game interface when the cheating tester simulates using the cheating during the game process, or it can be the cheating interface that appears on the game interface of the player using the cheating collected during the actual game process of the player, etc.

[0067] 102. Extract multiple associated word groups from each sample cheating interface image based on the accompanying relationship in the text content of the sample cheating interface image.

[0068] Among them, the associated word group refers to a word group composed of at least two words with an accompanying relationship in the text content of the sample cheating interface image.

[0069] Among them, the co-occurrence relationship refers to the association in terms of the positions where they appear in the text content, or the hierarchical layout in the text content, etc.

[0070] In some embodiments, in order to obtain accurate external keywords, the step of "extracting multiple associated phrases from each sample external interface image based on the co-occurrence relationship in the text content of the sample external interface image" may include the following operations:

[0071] Perform text recognition on each sample external interface image to obtain the text corresponding to each sample external interface image;

[0072] Determine from each text the word combinations composed of at least two words that co-occur;

[0073] Obtain multiple associated phrases in each text based on the word combinations in each text.

[0074] Among them, text recognition of the sample external interface image can use OCR (Optical Character Recognition) technology, which refers to the process of analyzing and recognizing the image file of text materials to obtain text and layout information. That is, the text in the image is recognized and returned in text form.

[0075] Specifically, text recognition processing of the sample external interface image by OCR technology may include the following steps: First, perform image preprocessing on the sample external interface image, including geometric transformation (perspective, distortion, rotation, etc.), distortion correction, deblurring, image enhancement, and light correction, etc.; then perform text detection on the preprocessed image to detect the location, range, and layout of the text; further, perform text recognition on the basis of text detection to recognize the text content, convert the text information in the image into text information, and obtain the text of the sample external interface image.

[0076] Among them, determining the word combinations composed of at least two words that co-occur from each text can be processed through the TextRank algorithm model. The TextRank algorithm is a graph-based ranking algorithm for text. By splitting the text into several constituent units (sentences / words), constructing a node connection graph, using the similarity between sentences (words) as the weight of the edge, calculating the TextRank value of sentences (words) through iterative loops, and finally extracting the sentence (word) combinations with high rankings to form a text summary.

[0077] Specifically, the text of the sample external interface image is segmented by the TextRank model to obtain individual sentences, and then each sentence is tokenized to obtain multiple words included in the text. Further, according to the position, range, and layout of each word in the text, the words that co-occur with the word among the adjacent words before and after the position of each word in the text are determined, and a word combination, that is, an associated phrase, is formed based on the word and the words that accompany it.

[0078] Furthermore, for the multiple word combinations extracted from the text of each sample external interface image, multiple associated phrases corresponding to each sample external interface image are obtained.

[0079] 103. Determine the external keyword score of each associated phrase according to the frequency of each associated phrase appearing in multiple sample external interface images.

[0080] Among them, the frequency of an associated phrase appearing in multiple sample external interface images is the proportion of the number of sample external interface images in which the associated phrase appears in all sample external interface images.

[0081] For example, the total number of sample external interface images can be 100, and the number of sample external interface images in which the associated phrase A appears can be 10. Then, it can be determined that the probability of the associated phrase A appearing in the sample external interface images is 10 / 100.

[0082] Among them, the external keyword score of an associated phrase refers to the score for evaluating whether the associated phrase belongs to an external keyword. The higher the external keyword score of an associated phrase, the greater the probability that the associated phrase is an external keyword.

[0083] In some embodiments, in order to make the external keywords obtained from the sample external interface images more representative, the step of "determining the external keyword score of each associated phrase according to the frequency of each associated phrase appearing in multiple sample external interface images" may include the following operations:

[0084] Determine the initial external keyword score corresponding to each associated phrase based on the frequency of each associated phrase appearing in multiple sample external interface images;

[0085] Determine the external keyword score of each associated phrase according to the universality of each associated phrase appearing in non-external interface images and the initial external keyword score.

[0086] Among them, non-external interface images refer to the acquired images of non-external interfaces.

[0087] First, according to the frequencies of each associated phrase in all sample external interface images, obtain the initial scores of the external keyword corresponding to each associated phrase. Then, according to the prevalence of each associated phrase in non-external interface images, suppress the initial scores of the external keywords of each associated word to obtain the final external keyword scores of each associated phrase.

[0088] In some embodiments, the step of "determining the external keyword score of each associated phrase according to the prevalence of each associated phrase in non-external interface images and the initial external keyword score" may include the following operations:

[0089] Obtain the set of non-external keyword groups corresponding to the non-external interface images;

[0090] Based on the prevalence of each associated phrase in the set of non-external keyword groups, determine the suppression coefficient of the external keyword score of each associated phrase;

[0091] Calculate the product of the initial external keyword score and the suppression coefficient of the external keyword score to obtain the external keyword score of the associated phrase.

[0092] Among them, the set of non-external keywords includes a plurality of non-external keywords extracted from the text of each non-external interface image, that is, the set of non-external keywords.

[0093] Then, the prevalence of each associated phrase in non-external interface images refers to the prevalence of each associated phrase in the set of non-external keywords.

[0094] In the embodiments of the present application, design the following algorithm formula to calculate the external keyword score of the associated phrase:

[0095]

[0096] Among them, c refers to the set of associated phrases extracted based on the sample external interface images, s refers to the set of non-external keyword groups extracted based on the non-external interface images, and a refers to any associated phrase in the set of associated words. c a / c calculates the frequency of each associated phrase appearing in multiple sample external interface images, calculates the prevalence of each associated phrase in the set of non-external keyword groups, that is, the suppression coefficient. Based on the product of the frequency and the suppression coefficient, obtain the external keyword score score of the associated phrase a.

[0097] 104. Determine the set of external keywords from all associated phrases based on the external keyword scores.

[0098] In some embodiments, in order to obtain a more representative set of external plug - in keywords, the step of "determining the set of external plug - in keywords from all associated phrases based on the external plug - in keyword scores" may include the following operations:

[0099] From all the associated phrases extracted from each sample external plug - in interface image, filter out the associated phrases whose external plug - in keyword scores are greater than a preset score to obtain the set of external plug - in keywords.

[0100] In the embodiments of the present application, a preset score can be set for filtering the final set of external plug - in keys.

[0101] Specifically, compare the external plug - in keyword scores of all the associated phrases extracted from each sample external plug - in interface image with the preset score, and filter out the associated phrases whose external plug - in keyword scores are greater than the preset score as the set of external plug - in keywords.

[0102] 105. Perform external plug - in detection on the candidate external plug - in interface images in the game interface to be detected based on the set of external plug - in keywords.

[0103] After obtaining the set of external plug - in keywords, external plug - in detection can be performed on the game interface based on the set of external plug - in keywords.

[0104] In some embodiments, in order to improve the accuracy of external plug - in detection, the step of "performing external plug - in detection on the candidate external plug - in interface images in the game interface to be detected based on the set of external plug - in keywords" may include the following operations:

[0105] Collect the game interface to be detected, identify non - game interface images from the game interface to be detected to obtain candidate external plug - in interface images;

[0106] Based on the accompanying relationship in the text content of the candidate external plug - in interface images, extract multiple candidate associated phrases from the candidate external plug - in interface images;

[0107] Match the multiple candidate associated phrases with the set of external plug - in keywords;

[0108] Determine the external plug - in detection result of the game interface to be detected based on the matching result.

[0109] Among them, the game interface to be detected can be the game interface of the player's terminal collected during the player's game process. By performing external plug - in detection on the game interface, it can be determined whether the player uses an external plug - in to play the game.

[0110] Among them, the non - game interface image refers to other interfaces displayed on the game interface that are irrelevant to the original game content. By identifying the game content of the game interface to be detected, an interface that does not include game content is obtained as the candidate external plug - in interface image.

[0111] Specifically, extracting candidate associated word groups from candidate external plug-in interface images can perform text recognition processing on the candidate external plug-in interface images through OCR technology to obtain the text of the candidate external plug-in interface images; further, the TextRank model is used to extract associated word groups from the text of the candidate external plug-in interface images to obtain multiple candidate associated word groups in the text of the candidate external plug-in interface images.

[0112] Further, multiple candidate associated word groups are matched with the external plug-in keyword groups to obtain a matching result, and then whether there is an external plug-in interface in the game interface to be detected is judged according to the matching result.

[0113] For example, the external plug-in keyword groups can include: the first external plug-in keyword group, the second external plug-in keyword group, the third external plug-in keyword group, the fourth external plug-in keyword group, etc., and multiple candidate associated word groups can include: the first candidate associated word group, the second candidate associated word group, the third candidate associated word group, etc. Then, the first candidate associated word group and the second candidate associated word group are respectively matched with the first external plug-in keyword group, the second external plug-in keyword group, the third external plug-in keyword group, and the fourth external plug-in keyword group to judge whether the candidate associated word group belongs to the external plug-in keyword.

[0114] In some embodiments, the step of "determining the external plug-in detection result of the game interface to be detected based on the matching result" may include the following operations:

[0115] If there is a candidate associated word group in the multiple candidate associated word groups that matches the external plug-in keyword group successfully, it is determined that the external plug-in detection result of the game interface to be detected is using an external plug-in;

[0116] If there is no candidate associated word group in the multiple candidate associated word groups that matches the external plug-in keyword group successfully, it is determined that the external plug-in detection result of the game interface to be detected is not using an external plug-in.

[0117] For example, by matching multiple candidate associated word groups with the external plug-in keyword groups respectively, if there is at least one associated word group in the multiple candidate associated word groups that matches the external plug-in keyword successfully, that is, at least one associated word group is the same as the external plug-in keyword, it can be determined that there is an external plug-in keyword group in the candidate external plug-in interface image, and then it is determined that there is a situation of using an external plug-in in the game interface to be detected.

[0118] Further, when it is detected that there is a situation of using an external plug-in in the game interface to be detected, the game player corresponding to the game interface to be detected is determined, and the game player is punished based on the external plug-in punishment strategy. The external plug-in punishment strategy can include: muting, banning, demoting, etc. Specifically, the corresponding external plug-in punishment strategy can be determined according to the influence degree of the game player using the external plug-in on other game players, so as to maintain the fairness of the game and reduce the game players using game external plug-ins.

[0119] For another example, by separately matching multiple candidate associated phrase groups with the external keyword group, if there is no associated phrase in the multiple candidate associated phrase groups that successfully matches the external keyword, that is, no associated phrase is the same as the external keyword, it can be determined that the candidate external interface image does not have the external keyword group, and further determine that there is no situation of using external software in the game interface to be detected.

[0120] In some embodiments, in order to automatically update the existing external keyword group library for more comprehensive external software detection, the method may further include the following steps:

[0121] If the external software detection result of the game interface to be detected indicates that no external software is used, add the candidate external interface image in the game interface to be detected to the set of potential external interface images;

[0122] Determine new external keyword groups based on the potential external interface images in the set of potential external interface images;

[0123] Update the external keyword group according to the new external keyword group, and perform external software detection on the candidate external interface images in the game interface to be detected based on the updated external keyword group.

[0124] Among them, the external software detection result of the game interface to be detected indicating that no external software is used means that there is no external keyword group in the candidate external interface image of the game interface to be detected.

[0125] In some embodiments, a new type of external software may appear. When a player uses the new external software to play the game, a new external interface may be displayed on the game interface, and this new external interface may be different from the sample external interface. At this time, if the external software detection is performed on this new external interface according to the external keywords extracted from the sample external interface image, this new external interface will be determined as an interface without using external software, resulting in an incorrect detection result.

[0126] Therefore, in this embodiment, a method for self-updating external keyword groups is designed, which may include: when performing external software detection on each candidate external interface image, for the candidate external interface images with a detection result of not using external software, collect and save them as potential external interface images to obtain a set of potential external interface images; then mine new external keyword groups from the set of potential external interface images.

[0127] In some embodiments, in order to automatically mine new external key groups, the step of "determining new external keyword groups based on the potential external interface images in the set of potential external interface images" may include the following operations:

[0128] Perform clustering processing on all potential external interface images in the set of potential external interface images to obtain multiple image subsets;

[0129] Filter out the image subsets that are not external interface images from multiple image subsets to obtain a target image subset;

[0130] Extract multiple potential associated phrase groups from each potential external interface image in the target image subset;

[0131] Determine the potential external keyword score of each potential associated phrase group according to the frequency of occurrence of each potential associated phrase group in the potential external interface images of the target image subset;

[0132] Determine a new external keyword group from multiple potential associated phrase groups based on the potential external keyword scores.

[0133] Specifically, when the number of potential external interface images in the potential external interface image set reaches a preset number, clustering processing can be performed on multiple potential external interface images. The clustering processing can divide multiple potential external interface images into multiple image subsets according to the image categories through a clustering algorithm, where each image subset includes potential external interface images belonging to the same category.

[0134] Among them, dividing multiple potential external interface images into multiple image subsets according to the image categories through a clustering algorithm can calculate the similarity between images, and obtain similar potential external interface images as an image subset.

[0135] In some embodiments, in order to obtain a real new external keyword group, it is necessary to filter out the image subsets that are not external interface images from multiple image subsets to obtain a target image subset.

[0136] Specifically, for each image subset, verify whether the images in the image subset are external interface images, so as to determine whether each image subset is an external interface image subset, and then filter out the image subsets of non-external interface images to obtain an external interface image subset, that is, the target image subset.

[0137] Among them, verifying the image subset can first train an independent phrase model through the text of the potential external interface images in the image subset, and then verify through a pre-designed external interface verification set to determine whether the image subset is an external interface image subset.

[0138] Furthermore, to extract potential associated phrase groups from each potential external interface image in the target image subset, text recognition processing can be performed on the potential external interface images through OCR technology to obtain the text of the potential external interface images; furthermore, through the TextRank model, potential associated phrase groups in the text of the potential external interface images are extracted to obtain multiple potential associated phrase groups in the text of the potential external interface images.

[0139] Among them, calculating the potential external plug-in keyword scores of each potential associated phrase can be based on the score calculation formula in this solution, that is:

[0140]

[0141] Finally, the potential associated phrases with potential external plug-in keyword scores greater than the preset score are used as the new external plug-in keyword groups.

[0142] Furthermore, update the external plug-in keyword group according to the new external plug-in keyword group, that is, add the new external plug-in keyword group to the existing external plug-in keyword group to obtain the updated external plug-in keyword group, and then perform external plug-in detection on the candidate external plug-in interface images in the game interface based on the updated external plug-in keyword group to improve the accuracy of external plug-in detection.

[0143] In some embodiments, if a target image subset is an external plug-in interface image subset, that is, the potential external plug-in interface images in the target image subset are external plug-in images, then the potential external plug-in interface images in the target image subset can be re-detected for external plug-ins based on the updated external plug-in keyword group to determine the game interface corresponding to the potential external plug-in interface image with the use of external plug-ins. Further, a certain degree of punishment can be imposed on the game player corresponding to the game interface. Through the re-detection method, the game interfaces with the use of external plug-ins that were missed in historical detections can be detected, and the game players can be punished to ensure the fairness of the game.

[0144] The embodiment of the present application discloses a method for detecting game external plug-ins. The method includes: obtaining a plurality of sample external plug-in interface images; extracting a plurality of associated phrases from each sample external plug-in interface image based on the accompanying relationship in the text content of the sample external plug-in interface images; determining the external plug-in keyword score of each associated phrase according to the frequency of occurrence of each associated phrase in the plurality of sample external plug-in interface images; determining an external plug-in keyword group from all the associated phrases based on the external plug-in keyword scores; and performing external plug-in detection on the candidate external plug-in interface images in the game interface to be detected based on the external plug-in keyword group, thereby improving the accuracy of game external plug-in detection.

[0145] According to the above-introduced content, the game external plug-in detection method of the present application will be further illustrated by examples below. Please refer to Figure 2 , Figure 2 which is a schematic flowchart of another game external plug-in detection method provided by the embodiment of the present application. Taking the application of this game external plug-in detection method to a server as an example, the specific process can be as follows:

[0146] 201. The server collects a plurality of external plug-in interface images.

[0147] Among them, the external interface image refers to the external interface that is superimposed on the game interface after using a game external program during the game operation.

[0148] In the embodiment of the present application, the external interface can be obtained by an external interface tester simulating the interfaces that are superimposed and displayed on the game interface when using various game external programs in the game. Among them, by simulating the external interfaces generated by using various game external programs, a more comprehensive external word spectrum library for external detection can be generated.

[0149] 202. The server extracts at least one accompanying word spectrum from each external interface image based on the text content and layout relationship of the external interface image.

[0150] Among them, the accompanying word spectrum refers to a phrase composed of at least two words that appear concomitantly in the text content of the external interface image.

[0151] Specifically, extracting the accompanying word spectrum from the external interface image may include: first, performing text recognition processing on the external interface image through OCR technology to obtain the text in the external interface image, and then, segmenting the text of the external interface image through the TextRank model to obtain individual sentences, and then performing word segmentation processing on each sentence to obtain multiple words included in the text; further, according to the positions, ranges and layouts of each word in the text, determining the words that appear concomitantly with the adjacent words before and after the position of each word in the text, and forming a word combination based on the word and the words that appear concomitantly with the word, that is, the accompanying word spectrum.

[0152] 203. The server calculates the external keyword score of each accompanying word spectrum, and screens out the external word spectrum from all the accompanying word spectra based on the external keyword score.

[0153] In the embodiment of the present application, the following algorithm formula is designed to calculate the external keyword score of the accompanying word spectrum:

[0154]

[0155] Among them, c is the external space, that is, the set of associated word groups extracted based on the external interface image, s is the normal sample space, that is, the set of non-external keyword groups extracted based on the non-external interface image, and a is the target word spectrum. score is the external keyword score of the target word spectrum.

[0156] In the embodiment of the present application, a second-order word spectrum is constructed through TextRank. TextRank finds keywords based on the entire text through sentence segmentation or character windows (word associations). The scenario of the embodiment of the present application can be the set of associated word groups extracted based on the external interface image and the set of non-external keyword groups extracted based on the non-external interface image, and an association relationship is constructed through the entire text or window accompaniment.

[0157] Use the scores of the calculated adjoint word spectra as the initial contribution coefficients of the keywords of each adjoint word spectrum in the external plug-in interface image. Since this score includes the suppression in normal samples, the extracted external plug-in keywords are more favorable and representative for external plug-ins.

[0158] Furthermore, for each adjoint word spectrum for which the external plug-in keyword score is calculated, the adjoint word spectrum whose external plug-in keyword score exceeds the preset score is used as the external plug-in word spectrum, that is, the keyword spectrum used for external plug-in detection.

[0159] 204. The server performs external plug-in detection on the interface image to be detected based on the external plug-in word spectrum, and obtains the external plug-in detection result of the interface image to be detected.

[0160] Specifically, extracting candidate associated word groups from the interface image to be detected can perform text recognition processing on the interface image to be detected through OCR technology to obtain the text of the interface image to be detected; further, perform adjoint word spectrum extraction on the text of the interface image to be detected through the TextRank model to obtain multiple adjoint word spectra to be detected in the text of the interface image to be detected.

[0161] Furthermore, match the multiple adjoint word spectra to be detected with the external plug-in word spectrum to obtain a matching result, and then judge whether the interface image to be detected uses an external plug-in according to the matching result.

[0162] Specifically, if the adjoint word spectrum to be detected matches the external plug-in word spectrum successfully, it means that the interface image to be detected uses an external plug-in; if the adjoint word spectrum to be detected fails to match the external plug-in word spectrum, it means that the interface image to be detected does not use an external plug-in.

[0163] In some embodiments, when performing external plug-in detection on the interface image to be detected, in addition to detecting whether there is an external plug-in in the interface image to be detected, potential external plug-in data can also be retained, and then new external plug-in word spectra can be mined through a clustering algorithm (verified and excluded as non-external plug-ins through a complete verification set), and then feedback to update the current model to achieve the purpose of self-update.

[0164] The specific implementation is as follows: For the interface image sample to be detected whose detection result is that there is no external plug-in, as a potential external plug-in sample, collect potential external plug-in samples; when the number of potential external plug-in samples accumulates to a certain level, cluster through a clustering algorithm (for example, through image similarity, but not the only one, any algorithm that can classify external plug-in interface images can be used) to obtain different external plug-in sample sets (the external plug-in interface images of the same class will be clustered together), and obtain at least one external plug-in sample cluster.

[0165] Then, for each cluster, an independent model similar to the graph model is trained and then verified and screened through a complete validation set. Specifically, for potential cheats, an independent word spectrum model is trained, verified and filtered through a complete validation set, real cheat samples are retained, merged into the full sample set to train the latest text model, and executed periodically, so as to achieve self-updating learning of the cheat word spectrum, extract new cheat word spectra from the retained cheat samples, update the cheat word spectrum based on the new cheat word spectra, and obtain the updated cheat word spectrum, thereby improving the accuracy of cheat detection.

[0166] An embodiment of the present application discloses a game cheat detection method, which includes: the server collects multiple cheat interface images, extracts at least one accompanying word spectrum from each cheat interface image based on the text content and layout relationship of the cheat interface images, calculates the cheat keyword score of each accompanying word spectrum, and screens out the cheat word spectrum from all accompanying word spectra based on the cheat keyword score, and performs cheat detection on the to-be-detected interface image based on the cheat word spectrum to obtain the cheat detection result of the to-be-detected interface image. In this way, the efficiency of game cheat detection in the game can be improved.

[0167] To facilitate better implementation of the game cheat detection method provided by the embodiments of the present application, the embodiments of the present application also provide a game cheat detection device based on the above game cheat detection method. The meanings of the nouns are the same as those in the above game cheat detection method, and the specific implementation details can refer to the descriptions in the method embodiments.

[0168] Please refer to Figure 3 , Figure 3 which is a structural block diagram of a game cheat detection device provided by an embodiment of the present application. The device includes:

[0169] An acquisition unit 301, configured to acquire multiple sample cheat interface images;

[0170] An extraction unit 302, configured to extract multiple associated word groups from each sample cheat interface image based on the accompanying relationship in the text content of the sample cheat interface image;

[0171] A first determination unit 303, configured to determine the cheat keyword score of each associated word group according to the frequency of occurrence of each associated word group in the multiple sample cheat interface images;

[0172] A second determination unit 304, configured to determine a cheat keyword group from all associated word groups based on the cheat keyword score;

[0173] A detection unit 305, configured to perform cheat detection on the candidate cheat interface images in the to-be-detected game interface based on the cheat keyword group.

[0174] In some embodiments, the first determination unit 303 may include:

[0175] A first determination subunit, configured to determine an initial score of an external plug-in keyword corresponding to each associated phrase based on the frequency of occurrence of each associated phrase in the multiple sample external plug-in interface images;

[0176] A second determination subunit, configured to determine an external plug-in keyword score of each associated phrase according to the universality of each associated phrase in the non-external plug-in interface images and the initial score of the external plug-in keyword.

[0177] In some embodiments, the second determination subunit may specifically be configured to:

[0178] Obtain a set of non-external plug-in keywords corresponding to the non-external plug-in interface images;

[0179] Determine an external plug-in keyword score suppression coefficient for each associated phrase based on the universality of each associated phrase in the set of non-external plug-in keywords;

[0180] Calculate a product value of the initial score of the external plug-in keyword and the external plug-in keyword score suppression coefficient to obtain the external plug-in keyword score of the associated phrase.

[0181] In some embodiments, the extraction unit 302 may include:

[0182] An identification unit, configured to perform text recognition on each sample external plug-in interface image to obtain text corresponding to each sample external plug-in interface image;

[0183] A third determination subunit, configured to determine, from the texts, a word combination composed of at least two words that appear concomitantly;

[0184] A fourth determination subunit, configured to obtain the multiple associated phrases in each text based on the word combinations in the texts.

[0185] In some embodiments, the second determination unit 304 may include:

[0186] A fifth determination subunit, configured to screen, from all the associated phrases extracted based on each sample external plug-in interface image, the associated phrases with an external plug-in keyword score greater than a preset score to obtain the set of external plug-in keywords.

[0187] In some embodiments, the detection unit 305 may include:

[0188] An acquisition subunit, configured to acquire a game interface to be detected, and identify a non-game interface image from the game interface to be detected to obtain the candidate external plug-in interface image;

[0189] The first extraction subunit is configured to extract a plurality of candidate associated phrase groups from the candidate external plug-in interface image based on the association relationship in the text content of the candidate external plug-in interface image;

[0190] The matching subunit is configured to match the plurality of candidate associated phrase groups with the external plug-in keyword group;

[0191] The sixth determination subunit is configured to determine the external plug-in detection result of the game interface to be detected based on the matching result.

[0192] In some embodiments, the sixth determination subunit may specifically be configured to:

[0193] If there is a candidate associated phrase group in the plurality of candidate associated phrase groups that matches the external plug-in keyword group, determine that the external plug-in detection result of the game interface to be detected is using an external plug-in;

[0194] If there is no candidate associated phrase group in the plurality of candidate associated phrase groups that matches the external plug-in keyword group, determine that the external plug-in detection result of the game interface to be detected is not using an external plug-in.

[0195] In some embodiments, the apparatus may further include:

[0196] The adding unit is configured to add the candidate external plug-in interface image in the game interface to be detected to the set of potential external plug-in interface images if the external plug-in detection result of the game interface to be detected indicates not using an external plug-in;

[0197] The third determination unit is configured to determine a new external plug-in keyword group according to the potential external plug-in interface images in the set of potential external plug-in interface images;

[0198] The updating unit is configured to update the external plug-in keyword group according to the new external plug-in keyword group, and perform external plug-in detection on the candidate external plug-in interface images in the game interface to be detected based on the updated external plug-in keyword group.

[0199] In some embodiments, the third determination unit may include:

[0200] The processing subunit is configured to perform clustering processing on all the potential external plug-in interface images in the set of potential external plug-in interface images to obtain a plurality of image subsets, where each image subset includes potential external plug-in interface images belonging to the same category;

[0201] The eliminating subunit is configured to eliminate the image subsets of non-external plug-in interface images from the plurality of image subsets to obtain a target image subset;

[0202] The second extraction subunit is configured to extract a plurality of potential associated phrase groups from each potential external plug-in interface image in the target image subset;

[0203] A seventh determination subunit, configured to determine a potential plug-in keyword score for each potential associated phrase according to the frequency of occurrence of each potential associated phrase in the potential plug-in interface images of the target image subset;

[0204] An eighth determination subunit, configured to determine the new plug-in keyword group from multiple potential associated phrases based on the potential plug-in keyword scores.

[0205] An embodiment of the present application discloses a game plug-in detection device. The acquisition unit 301 acquires a plurality of sample plug-in interface images; the extraction unit 302 extracts a plurality of associated phrases from each sample plug-in interface image based on the accompanying relationship in the text content of the sample plug-in interface images; the first determination unit 303 determines a plug-in keyword score for each associated phrase according to the frequency of occurrence of each associated phrase in the plurality of sample plug-in interface images; the second determination unit 304 determines a plug-in keyword group from all the associated phrases based on the plug-in keyword scores; the detection unit 305 performs plug-in detection on the candidate plug-in interface images in the game interface to be detected. In this way, the accuracy of game plug-in detection can be improved.

[0206] Correspondingly, an embodiment of the present application further provides a computer device, and the computer device may be a terminal. As Figure 4 shown, Figure 4 is a schematic structural diagram of the computer device provided by the embodiment of the present application. The computer device 500 includes a processor 501 having one or more processing cores, a memory 502 having one or more computer-readable storage media, and a computer program stored in the memory 502 and executable on the processor. Among them, the processor 501 is electrically connected to the memory 502. Those skilled in the art can understand that the structural diagram of the computer device shown in the figure does not constitute a limitation on the computer device, and it may include more or fewer components than shown in the figure, or combine some components, or have different component arrangements.

[0207] The processor 501 is the control center of the computer device 500, connects various parts of the entire computer device 500 through various interfaces and lines, and by running or loading software programs and / or modules stored in the memory 502, and calling data stored in the memory 502, executes various functions of the computer device 500 and processes data, thereby performing overall monitoring of the computer device 500.

[0208] In the embodiment of the present application, the processor 501 in the computer device 500 will load the instructions corresponding to the processes of one or more application programs into the memory 502 according to the following steps, and the processor 501 will run the application programs stored in the memory 502 to implement various functions:

[0209] Obtain multiple sample external cheat interface images;

[0210] Extract multiple associated phrase groups from each sample external cheat interface image based on the co-occurrence relationship in the text content of the sample external cheat interface images;

[0211] Determine the external cheat keyword score for each associated phrase group according to the frequency of occurrence of each associated phrase group in multiple sample external cheat interface images;

[0212] Determine the external cheat keyword group from all associated phrase groups based on the external cheat keyword scores;

[0213] Perform external cheat detection on the candidate external cheat interface images in the game interface to be detected based on the external cheat keyword group.

[0214] In some embodiments, determining the external cheat keyword score for each associated phrase group according to the frequency of occurrence of each associated phrase group in multiple sample external cheat interface images includes:

[0215] Determine the initial external cheat keyword score corresponding to each associated phrase group based on the frequency of occurrence of each associated phrase group in multiple sample external cheat interface images;

[0216] Determine the external cheat keyword score for each associated phrase group according to the universality of each associated phrase group in non-external cheat interface images and the initial external cheat keyword score.

[0217] In some embodiments, determining the external cheat keyword score for each associated phrase group according to the universality of each associated phrase group in non-external cheat interface images and the initial external cheat keyword score includes:

[0218] Obtain the non-external cheat keyword group set corresponding to the non-external cheat interface images;

[0219] Determine the external cheat keyword score suppression coefficient for each associated phrase group based on the universality of each associated phrase group in the non-external cheat keyword group set;

[0220] Calculate the product of the initial external cheat keyword score and the external cheat keyword score suppression coefficient to obtain the external cheat keyword score of the associated phrase group.

[0221] In some embodiments, extracting multiple associated phrase groups from each sample external cheat interface image based on the co-occurrence relationship in the text content of the sample external cheat interface images includes:

[0222] Perform text recognition on each sample external cheat interface image to obtain the text corresponding to each sample external cheat interface image;

[0223] Determine the word combinations composed of at least two words that co-occur from each text;

[0224] Obtain multiple associated phrase groups in each text based on the combination of words in each text.

[0225] In some embodiments, determining an external plug-in keyword group from all the associated phrase groups based on the external plug-in keyword score includes:

[0226] Screen out the associated phrase groups with an external plug-in keyword score greater than a preset score from all the associated phrase groups extracted from each sample external plug-in interface image to obtain the external plug-in keyword group.

[0227] In some embodiments, performing external plug-in detection on the candidate external plug-in interface image in the game interface to be detected based on the external plug-in keyword group includes:

[0228] Collect the game interface to be detected, and identify non-game interface images from the game interface to be detected to obtain the candidate external plug-in interface image;

[0229] Extract multiple candidate associated phrase groups from the candidate external plug-in interface image based on the accompanying relationship in the text content of the candidate external plug-in interface image;

[0230] Match the multiple candidate associated phrase groups with the external plug-in keyword group;

[0231] Determine the external plug-in detection result of the game interface to be detected based on the matching result.

[0232] In some embodiments, determining the external plug-in detection result of the game interface to be detected based on the matching result includes:

[0233] If there is a candidate associated phrase group that matches successfully with the external plug-in keyword group among the multiple candidate associated phrase groups, determine that the external plug-in detection result of the game interface to be detected is using an external plug-in;

[0234] If there is no candidate associated phrase group that matches successfully with the external plug-in keyword group among the multiple candidate associated phrase groups, determine that the external plug-in detection result of the game interface to be detected is not using an external plug-in.

[0235] In some embodiments, the method further includes:

[0236] If the external plug-in detection result of the game interface to be detected indicates not using an external plug-in, add the candidate external plug-in interface image in the game interface to be detected to the set of potential external plug-in interface images;

[0237] Determine a new external plug-in keyword group according to the potential external plug-in interface images in the set of potential external plug-in interface images;

[0238] Update the external plug-in keyword group according to the new external plug-in keyword group, and perform external plug-in detection on the candidate external plug-in interface image in the game interface to be detected based on the updated external plug-in keyword group.

[0239] In some embodiments, determining a new set of cheating keywords based on potential cheating interface images in a set of potential cheating interface images includes:

[0240] Performing clustering processing on all potential cheating interface images in the set of potential cheating interface images to obtain a plurality of image subsets, where each image subset includes potential cheating interface images belonging to the same category;

[0241] Removing image subsets that are not cheating interface images from the plurality of image subsets to obtain target image subsets;

[0242] Extracting a plurality of potential associated phrase groups from each potential cheating interface image in the target image subsets;

[0243] Determining the potential cheating keyword score of each potential associated phrase group according to the frequency of occurrence of each potential associated phrase group in the potential cheating interface images of the target image subsets;

[0244] Determining a new set of cheating keywords from the plurality of potential associated phrase groups based on the potential cheating keyword scores.

[0245] This solution improves the accuracy of game cheating detection by obtaining a plurality of sample cheating interface images; extracting a plurality of associated phrase groups from each sample cheating interface image based on the accompanying relationship in the text content of the sample cheating interface images; determining the cheating keyword score of each associated phrase group according to the frequency of occurrence of each associated phrase group in the plurality of sample cheating interface images; determining a set of cheating keywords from all the associated phrase groups based on the cheating keyword scores; and performing cheating detection on candidate cheating interface images in the game interface to be detected based on the set of cheating keywords.

[0246] For the specific implementation of each of the above operations, reference may be made to the previous embodiments and will not be elaborated herein.

[0247] Optionally, as Figure 4 shown, the computer device 500 further includes: a touch display screen 503, a radio frequency circuit 504, an audio circuit 505, an input unit 506, and a power supply 507. Among them, the processor 501 is electrically connected to the touch display screen 503, the radio frequency circuit 504, the audio circuit 505, the input unit 506, and the power supply 507 respectively. Those skilled in the art can understand that Figure 4 the computer device structure shown in

[0248] The touch display screen 503 can be used to display a graphical user interface and receive operation instructions generated by a user's interaction with the graphical user interface. The touch display screen 503 may include a display panel and a touch panel. Among them, the display panel can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the computer device. These graphical user interfaces can be composed of graphics, guiding information, icons, videos, and any combination thereof. Optionally, the display panel can be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), etc. The touch panel can be used to collect touch operations of the user on or near it (such as operations of the user using a finger, a stylus, or any suitable object or accessory on or near the touch panel), and generate corresponding operation instructions, and the operation instructions execute corresponding programs. Optionally, the touch panel can include two parts: a touch detection device and a touch controller. Among them, the touch detection device detects the touch position of the user and detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into touch point coordinates, and then sends it to the processor 501, and can receive and execute commands sent by the processor 501. The touch panel can cover the display panel. After the touch panel detects a touch operation on or near it, it is transmitted to the processor 501 to determine the type of touch event. Subsequently, the processor 501 provides a corresponding visual output on the display panel according to the type of touch event. In the embodiments of the present application, the touch panel and the display panel can be integrated into the touch display screen 503 to implement input and output functions. However, in some embodiments, the touch panel and the touch panel can be implemented as two independent components to implement input and output functions. That is, the touch display screen 503 can also be used as a part of the input unit 506 to implement the input function.

[0249] The radio frequency circuit 504 can be used to transmit and receive radio frequency signals to establish wireless communication with a network device or other computer devices through wireless communication, and transmit and receive signals with the network device or other computer devices.

[0250] The audio circuit 505 can be used to provide an audio interface between the user and the computer device through a speaker and a microphone. The audio circuit 505 can transmit the electrical signal converted from the received audio data to the speaker, and the speaker converts it into a sound signal for output; on the other hand, the microphone converts the collected sound signal into an electrical signal, which is received by the audio circuit 505 and then converted into audio data. After the audio data is output to the processor 501 for processing, it is transmitted through the radio frequency circuit 504 to, for example, another computer device, or the audio data is output to the memory 502 for further processing. The audio circuit 505 may also include an earphone jack to provide communication between a peripheral earphone and the computer device.

[0251] The input unit 506 can be used to receive input digital, character information or user feature information (such as fingerprints, irises, facial information, etc.), and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function controls.

[0252] The power supply 507 is used to supply power to each component of the computer device 500. Optionally, the power supply 507 can be logically connected to the processor 501 through a power management system, so as to implement functions such as management of charging, discharging, and power consumption management through the power management system. The power supply 507 can also include any components such as one or more DC or AC power supplies, a recharge system, a power failure detection circuit, a power converter or inverter, and a power status indicator.

[0253] Although Figure 4 not shown in the figure, the computer device 500 may also include a camera, a sensor, a Wi-Fi module, a Bluetooth module, etc., which will not be elaborated here.

[0254] In the above embodiments, the descriptions of the respective embodiments have their own focuses. For parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0255] As can be seen from the above, the computer device provided in this embodiment obtains a plurality of sample external interface images; extracts a plurality of associated phrases from each sample external interface image based on the accompanying relationship in the text content of the sample external interface images; determines the external keyword score of each associated phrase according to the frequency of occurrence of each associated phrase in the plurality of sample external interface images; determines an external keyword group from all the associated phrases based on the external keyword score; and performs external detection on the candidate external interface images in the game interface to be detected based on the external keyword group.

[0256] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions, or by controlling relevant hardware through instructions. The instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0257] Therefore, an embodiment of the present application provides a computer-readable storage medium, in which multiple computer programs are stored. The computer programs can be loaded by a processor to execute the steps in any one of the game external detection methods provided by the embodiments of the present application. For example, the computer program can execute the following steps:

[0258] Obtain a plurality of sample external interface images;

[0259] Extract multiple associated phrase groups from each sample cheating interface image based on the co-occurrence relationship in the text content of the sample cheating interface image;

[0260] Determine the cheating keyword score for each associated phrase group according to the frequency of occurrence of each associated phrase group in multiple sample cheating interface images;

[0261] Determine the cheating keyword group from all associated phrase groups based on the cheating keyword scores;

[0262] Perform cheating detection on the candidate cheating interface images in the game interface to be detected based on the cheating keyword group.

[0263] In some embodiments, determining the cheating keyword score for each associated phrase group according to the frequency of occurrence of each associated phrase group in multiple sample cheating interface images includes:

[0264] Determine the initial cheating keyword score corresponding to each associated phrase group based on the frequency of occurrence of each associated phrase group in multiple sample cheating interface images;

[0265] Determine the cheating keyword score for each associated phrase group according to the prevalence of each associated phrase group in non-cheating interface images and the initial cheating keyword score.

[0266] In some embodiments, determining the cheating keyword score for each associated phrase group according to the prevalence of each associated phrase group in non-cheating interface images and the initial cheating keyword score includes:

[0267] Obtain the non-cheating keyword group set corresponding to the non-cheating interface images;

[0268] Determine the cheating keyword score suppression coefficient for each associated phrase group based on the prevalence of each associated phrase group in the non-cheating keyword group set;

[0269] Calculate the product of the initial cheating keyword score and the cheating keyword score suppression coefficient to obtain the cheating keyword score of the associated phrase group.

[0270] In some embodiments, extracting multiple associated phrase groups from each sample cheating interface image based on the co-occurrence relationship in the text content of the sample cheating interface image includes:

[0271] Perform text recognition on each sample cheating interface image to obtain the text corresponding to each sample cheating interface image;

[0272] Determine from each text the word combinations composed of at least two words that co-occur;

[0273] Obtain multiple associated phrase groups in each text based on the word combinations in each text.

[0274] In some embodiments, determining the external plug-in keyword group from all associated phrase groups based on the external plug-in keyword score includes:

[0275] Filtering out the associated phrase groups with an external plug-in keyword score greater than a preset score from all the associated phrase groups extracted from each sample external plug-in interface image to obtain the external plug-in keyword group.

[0276] In some embodiments, performing external plug-in detection on the candidate external plug-in interface images in the game interface to be detected based on the external plug-in keyword group includes:

[0277] Collecting the game interface to be detected, and identifying non-game interface images from the game interface to be detected to obtain candidate external plug-in interface images;

[0278] Extracting a plurality of candidate associated phrase groups from the candidate external plug-in interface images based on the accompanying relationship in the text content of the candidate external plug-in interface images;

[0279] Matching the plurality of candidate associated phrase groups with the external plug-in keyword group;

[0280] Determining the external plug-in detection result of the game interface to be detected based on the matching result.

[0281] In some embodiments, determining the external plug-in detection result of the game interface to be detected based on the matching result includes:

[0282] If there is a candidate associated phrase group that matches successfully with the external plug-in keyword group among the plurality of candidate associated phrase groups, determining that the external plug-in detection result of the game interface to be detected is using an external plug-in;

[0283] If there is no candidate associated phrase group that matches successfully with the external plug-in keyword group among the plurality of candidate associated phrase groups, determining that the external plug-in detection result of the game interface to be detected is not using an external plug-in.

[0284] In some embodiments, the method further includes:

[0285] If the external plug-in detection result of the game interface to be detected indicates not using an external plug-in, adding the candidate external plug-in interface images in the game interface to be detected to the set of potential external plug-in interface images;

[0286] Determining a new external plug-in keyword group according to the potential external plug-in interface images in the set of potential external plug-in interface images;

[0287] Updating the external plug-in keyword group according to the new external plug-in keyword group, and performing external plug-in detection on the candidate external plug-in interface images in the game interface to be detected based on the updated external plug-in keyword group.

[0288] In some embodiments, determining a new external plug-in keyword group according to the potential external plug-in interface images in the set of potential external plug-in interface images includes:

[0289] Cluster all potential cheating interface images in the set of potential cheating interface images to obtain multiple image subsets, where each image subset includes potential cheating interface images belonging to the same category;

[0290] Remove the image subsets that are not cheating interface images from the multiple image subsets to obtain the target image subsets;

[0291] Extract multiple potential associated phrases from each potential cheating interface image in the target image subsets;

[0292] Determine the potential cheating keyword scores for each potential associated phrase according to the frequency of occurrence of each potential associated phrase in the potential cheating interface images of the target image subsets;

[0293] Determine a new set of cheating keywords from the multiple potential associated phrases based on the potential cheating keyword scores.

[0294] This solution improves the accuracy of game cheating detection by obtaining multiple sample cheating interface images; extracting multiple associated phrases from each sample cheating interface image based on the co-occurrence relationship in the text content of the sample cheating interface images; determining the cheating keyword scores for each associated phrase according to the frequency of occurrence of each associated phrase in the multiple sample cheating interface images; determining a set of cheating keywords from all the associated phrases based on the cheating keyword scores; and performing cheating detection on the candidate cheating interface images in the game interface to be detected using the set of cheating keywords.

[0295] For the specific implementation of each of the above operations, please refer to the previous embodiments and will not be elaborated here.

[0296] Among them, the storage medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, etc.

[0297] Since the computer program stored in the storage medium can execute the steps in any of the game cheating detection methods provided in the embodiments of the present application, the beneficial effects achievable by any of the game cheating detection methods provided in the embodiments of the present application can be realized. For details, please refer to the previous embodiments and will not be elaborated here.

[0298] The above has introduced in detail a game cheating detection method, device, storage medium and computer device provided by the embodiments of the present application. Specific examples are used in this article to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for detecting game cheats, characterized in that, The method includes: Obtaining a plurality of sample external cheat interface images; Extracting a plurality of associated phrase groups from each of the sample external cheat interface images based on the co-occurrence relationships among the text contents of the sample external cheat interface images; Determining an initial score of an external cheat keyword corresponding to each associated phrase group based on the frequency of occurrence of each associated phrase group in the plurality of sample external cheat interface images; Determining the score of an external cheat keyword for each associated phrase group according to the universality of each associated phrase group in non-external cheat interface images and the initial score of the external cheat keyword; Determining an external cheat keyword group from all the associated phrase groups based on the scores of the external cheat keywords; Performing external cheat detection on candidate external cheat interface images in a game interface to be detected based on the external cheat keyword group; If the external cheat detection result of the game interface to be detected indicates that no external cheat is used, adding the candidate external cheat interface images in the game interface to be detected to a set of potential external cheat interface images; Determining a new external cheat keyword group according to the potential external cheat interface images in the set of potential external cheat interface images; Updating the external cheat keyword group according to the new external cheat keyword group, and performing external cheat detection on candidate external cheat interface images in the game interface to be detected based on the updated external cheat keyword group.

2. The method according to claim 1, wherein The determining the score of an external cheat keyword for each associated phrase group according to the universality of each associated phrase group in non-external cheat interface images and the initial score of the external cheat keyword includes: Obtaining a set of non-external cheat keyword groups corresponding to the non-external cheat interface images; Determining an inhibition coefficient of the score of an external cheat keyword for each associated phrase group based on the universality of each associated phrase group in the set of non-external cheat keyword groups; Calculating a product value of the initial score of the external cheat keyword and the inhibition coefficient of the score of the external cheat keyword to obtain the score of the external cheat keyword for the associated phrase group.

3. The method according to claim 1, wherein The extracting a plurality of associated phrase groups from each of the sample external cheat interface images based on the co-occurrence relationships among the text contents of the sample external cheat interface images includes: Performing text recognition on each of the sample external cheat interface images to obtain the text corresponding to each of the sample external cheat interface images; Determining, from each text, a combination of words consisting of at least two words that co-occur; Obtaining the plurality of associated phrase groups in each text based on the combinations of words in each text.

4. The method according to claim 1, wherein The determining an external cheat keyword group from all the associated phrase groups based on the scores of the external cheat keywords includes: Screening, from all the associated phrase groups extracted from the sample external cheat interface images, the associated phrase groups with scores of external cheat keywords greater than a preset score to obtain the external cheat keyword group.

5. The method according to claim 1, wherein The performing external cheat detection on candidate external cheat interface images in a game interface to be detected based on the external cheat keyword group includes: Collecting a game interface to be detected, and identifying non-game interface images from the game interface to be detected to obtain the candidate external cheat interface images; Extracting a plurality of candidate associated phrase groups from the candidate external cheat interface images based on the co-occurrence relationships among the text contents of the candidate external cheat interface images; Matching the plurality of candidate associated phrase groups with the external cheat keyword group; Determining the external cheat detection result of the game interface to be detected based on the matching result.

6. The method according to claim 5, characterized in that Determining the cheating detection result of the game interface to be detected based on the matching result includes: If there is a candidate associated phrase in the multiple candidate associated phrases that matches the cheating keyword phrase, determine that the cheating detection result of the game interface to be detected is using cheating software; If there is no candidate associated phrase in the multiple candidate associated phrases that matches the cheating keyword phrase, determine that the cheating detection result of the game interface to be detected is not using cheating software.

7. The method according to claim 1, characterized in that, The determining a new cheating keyword phrase according to the potential cheating interface images in the potential cheating interface image set includes: Performing clustering processing on all the potential cheating interface images in the potential cheating interface image set to obtain a plurality of image subsets, where each image subset includes potential cheating interface images belonging to the same category; Removing the image subsets of non-cheating interface images from the plurality of image subsets to obtain a target image subset; Extracting a plurality of potential associated phrases from each potential cheating interface image in the target image subset; Determining the potential cheating keyword score of each potential associated phrase according to the frequency of occurrence of each potential associated phrase in the potential cheating interface images of the target image subset; Determining the new cheating keyword phrase from the multiple potential associated phrases based on the potential cheating keyword scores.

8. A game cheating detection device, characterized in that, The device includes: An obtaining unit, configured to obtain a plurality of sample cheating interface images; An extracting unit, configured to extract a plurality of associated phrases from each sample cheating interface image based on the accompanying relationship in the text content of the sample cheating interface image; A first determining unit, configured to determine the initial cheating keyword score corresponding to each associated phrase according to the frequency of occurrence of each associated phrase in the plurality of sample cheating interface images; and determine the cheating keyword score of each associated phrase according to the universality of each associated phrase in the non-cheating interface images and the initial cheating keyword score; A second determining unit, configured to determine a cheating keyword phrase from all the associated phrases based on the cheating keyword scores; A detecting unit, configured to perform cheating detection on the candidate cheating interface images in the game interface to be detected based on the cheating keyword phrase; An adding unit, configured to add the candidate cheating interface images in the game interface to be detected to the potential cheating interface image set if the cheating detection result of the game interface to be detected indicates that no cheating software is used; A third determining unit, configured to determine a new cheating keyword phrase according to the potential cheating interface images in the potential cheating interface image set; An updating unit, configured to update the cheating keyword phrase according to the new cheating keyword phrase, and perform cheating detection on the candidate cheating interface images in the game interface to be detected based on the updated cheating keyword phrase.

9. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored on the memory and running on the processor, where the processor implements the game cheating detection method according to any one of claims 1 to 7 when executing the program.

10. A storage medium, characterized in that, The storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by the processor to execute the game cheating detection method according to any one of claims 1 to 7.

Images