Game interface data recognition method and system

Abstract

The present invention belongs to the field of data recognition and discloses a game interface data recognition method and system, including: S1, taking a screenshot of the game interface using an adaptive screenshot interval to obtain a screenshot image; S2, identifying the screenshot image and calculating the first similarity between the screenshot image and a pre-stored strategy image; S3, judging whether the maximum first similarity is greater than a set similarity threshold, and if so, obtaining the pre-stored strategy image corresponding to the maximum first similarity; S4, obtaining a corresponding list of game strategies based on the pre-stored strategy image corresponding to the maximum first similarity; S5, after detecting that a user clicks on a floating ball displayed at the edge of the game interface, closing the floating ball and displaying the list of game strategies in the form of a pop-up window; after detecting that the user clicks on the zoom button of the pop-up window, closing the pop-up window and redisplaying the floating ball at the edge of the game interface. The present invention can conveniently and quickly provide users with accurate game strategies.

Description

Technical Field

[0001] The present invention relates to the field of data recognition, and in particular to a method and system for recognizing game interface data. Background Art

[0002] With the development of smartphones, more and more people are playing games on them. For games with complex plots, such as RPG games, users often need to read a guide while playing. However, due to the small screen size of smartphones, the guide display interface and the game interface cannot be displayed simultaneously on the same screen. Therefore, users are usually forced to switch back and forth between the guide display interface and the game interface to refer to the guide and play the game.

[0003] To address this technical issue, some existing technologies have adopted floating ball technology to display game strategies. Specifically, the user first searches for the game strategy they need, then shrinks the strategy display interface into a floating ball. During the game operation, the user clicks the floating ball to pop up the strategy display interface. After viewing the strategy, the user clicks the shrink button on the display interface to shrink the game interface back to the floating ball. However, this also brings a problem, that is, the user still needs to search for game strategies on their own, and the search results are relatively dependent on the keywords entered by the user. As a result, strategies that can solve the problem the user wants to solve are not correctly searched out due to inaccurate keyword input. Summary of the Invention

[0004] The purpose of the present invention is to disclose a game interface data recognition method and system to solve the technical problems raised in the background technology.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] In a first aspect, the present invention provides a method for identifying game interface data, comprising:

[0007] S1, taking a screenshot of the game interface using an adaptive screenshot interval to obtain a screenshot image;

[0008] S2, identifying the screenshot image and calculating a first similarity between the screenshot image and a pre-stored strategy image;

[0009] S3, determining whether the maximum first similarity is greater than a set similarity threshold, and if so, obtaining a pre-stored strategy image corresponding to the maximum first similarity;

[0010] S4, obtaining a corresponding game strategy list according to the pre-stored strategy image corresponding to the largest first similarity;

[0011] S5, after detecting that the user clicks on the floating ball displayed at the edge of the game interface, close the floating ball and display the list of game strategies in the form of a pop-up window; after detecting that the user clicks on the shrink button of the pop-up window, close the pop-up window and display the floating ball again at the edge of the game interface.

[0012] Preferably, the edges of the game interface include an upper edge, a lower edge, a left edge and a right edge.

[0013] Preferably, the floating ball displayed at the edge of the game interface includes a floating ball displayed at the upper edge.

[0014] Preferably, taking screenshots of the game interface using an adaptive screenshot interval includes:

[0015] After the countdown corresponding to the k-1th screenshot interval ends, take a screenshot of the game interface. After the screenshot is completed, calculate the kth screenshot interval t k , t k The calculation formula is:

[0016]

[0017] sim k-1,k-2 represents the second similarity between the screenshot images obtained after the countdown corresponding to the k-1th and k-2th screenshot intervals ends, simtop represents the upper limit of the second similarity, ncp represents the preset value, and nclic represents the time interval [t end,k-1 -T,t end,k-1 ], the number of times the user clicks on the floating ball, t end,k-1 Indicates the time when the countdown corresponding to the k-1th screenshot interval ends, and occrat indicates the time at which the countdown ends. end,k-1 The CPU occupancy rate of the device running the game. occtop indicates the upper limit of the occupancy rate. bs Indicates the set duration, α1, α2, and α3 represent the weight of similarity, the weight of the number of clicks, and the weight of CPU occupancy respectively; T represents the length of the time interval;

[0018] In calculating t k After that, a countdown is performed, and the countdown duration is the k-th screenshot interval t k .

[0019] Preferably, the pre-stored strategy images are screenshot images of the game included in the game strategy.

[0020] Preferably, calculating the first similarity between the screenshot image and the pre-stored strategy image includes:

[0021] For the screenshot image P1 and the pre-stored strategy image P2, the calculation process of the first similarity between P1 and P2 is as follows:

[0022] If the resolution of P1 is greater than that of P2, P1 is reduced to the same resolution as P2 to obtain image P1b, and the first similarity between P1b and P2 is calculated;

[0023] If the resolution of P1 is smaller than that of P2, P2 is reduced to the same resolution as P1 to obtain image P2b, and a first similarity between P1 and P2b is calculated;

[0024] If the resolution of P1 is the same as the resolution of P2, the first similarity between P1 and P2 is directly calculated.

[0025] Preferably, for two screenshot images Pb and strategy images Pc with the same resolution, the first similarity between Pb and Pc is calculated as follows:

[0026] Calculate Pb and Pc using the preset mask image to obtain images Pb1 and Pc1;

[0027] Calculate the number of sampling areas;

[0028] A first similarity between Pb1 and Pc1 is calculated based on the number of sampling regions.

[0029] Preferably, obtaining a list of corresponding game strategies according to the pre-stored strategy image corresponding to the largest first similarity includes:

[0030] Each game guide contains at least one guide image;

[0031] A list of game guides is generated based on hyperlink titles of all game guides containing pre-stored guide images corresponding to the largest first similarities.

[0032] Preferably, generating a list of game guides based on the hyperlink titles of all game guides containing pre-stored guide images corresponding to the largest first similarity comprises:

[0033] All hyperlink titles of game guides containing the pre-stored guide image corresponding to the largest first similarity are sorted in descending order according to the release time of the game guides to obtain a list of game guides.

[0034] In a second aspect, the present invention provides a game interface data recognition system, comprising a screenshot module, a recognition module, a judgment module, an acquisition module, and a detection module;

[0035] The screenshot module is used to take screenshots of the game interface using an adaptive screenshot interval to obtain screenshot images;

[0036] The recognition module is used to recognize the screenshot image and calculate the first similarity between the screenshot image and the pre-stored strategy image;

[0037] The judging module is used to judge whether the maximum first similarity is greater than a set similarity threshold, and if so, obtain the pre-stored strategy image corresponding to the maximum first similarity;

[0038] The acquisition module is used to acquire a list of corresponding game strategies according to the pre-stored strategy image corresponding to the largest first similarity;

[0039] The detection module is used to close the floating ball displayed on the edge of the game interface after detecting that the user clicks on the floating ball, and display the list of game strategies in the form of a pop-up window; after detecting that the user clicks on the shrink button of the pop-up window, close the pop-up window and redisplay the floating ball on the edge of the game interface.

[0040] Beneficial effects:

[0041] In the process of data recognition of the game interface on the smart phone, the present invention obtains a list of corresponding game strategies by calculating the first similarity between the screenshot image and the strategy image contained in the game strategy, thereby eliminating the need for the user to manually search for the game strategy, which is conducive to providing the user with more accurate game strategies more conveniently and quickly. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 A schematic diagram of a game interface data recognition method of the present invention.

[0044] Figure 2 A schematic diagram of a game interface data recognition system of the present invention. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0046] In a first aspect, the present invention provides a method for identifying game interface data, comprising:

[0047] S1, taking a screenshot of the game interface using an adaptive screenshot interval to obtain a screenshot image;

[0048] S2, identifying the screenshot image and calculating a first similarity between the screenshot image and a pre-stored strategy image;

[0049] S3, determining whether the maximum first similarity is greater than a set similarity threshold, and if so, obtaining a pre-stored strategy image corresponding to the maximum first similarity;

[0050] S4, obtaining a corresponding game strategy list according to the pre-stored strategy image corresponding to the largest first similarity;

[0051] S5, after detecting that the user clicks on the floating ball displayed at the edge of the game interface, close the floating ball and display the list of game strategies in the form of a pop-up window; after detecting that the user clicks on the shrink button of the pop-up window, close the pop-up window and display the floating ball again at the edge of the game interface.

[0052] In the process of data recognition of the game interface on the smartphone, a list of corresponding game guides is obtained by calculating the first similarity between the screenshot image and the guide image contained in the game guide, thereby eliminating the need for users to manually search for game guides, which is conducive to providing users with more accurate game guides more conveniently and quickly.

[0053] Specifically, the similarity threshold may be set to 0.8.

[0054] Specifically, the size of the pop-up window is smaller than the size of the game interface, and a list of game strategies is displayed in the center of the pop-up window. When the user clicks the name of the game strategy in the list, the detailed game strategy will be displayed in the pop-up window.

[0055] Preferably, the edges of the game interface include an upper edge, a lower edge, a left edge and a right edge.

[0056] Specifically, any one of the four edges may be displayed on the floating ball.

[0057] Preferably, the floating ball displayed at the edge of the game interface includes a floating ball displayed at the upper edge.

[0058] Since both sides of the lower edge are usually close to the virtual buttons for operating the game characters, and the left edge or the right edge usually displays information about the ongoing task, the floating ball is displayed on the upper edge by default.

[0059] The floating ball can adopt a translucent design to reduce the degree of obstruction of content on the game interface.

[0060] Preferably, taking screenshots of the game interface using an adaptive screenshot interval includes:

[0061] After the countdown corresponding to the k-1th screenshot interval ends, take a screenshot of the game interface. After the screenshot is completed, calculate the kth screenshot interval t k , t k The calculation formula is:

[0062]

[0063] sim k-1,k-2 represents the second similarity between the screenshot images obtained after the countdown corresponding to the k-1th and k-2th screenshot intervals ends, simtop represents the upper limit of the second similarity, ncp represents the preset value, and nclic represents the time interval [t end,k-1 -T,t end,k-1 ], the number of times the user clicks on the floating ball, t end,k-1 Indicates the time when the countdown corresponding to the k-1th screenshot interval ends, and occrat indicates the time at which the countdown ends. end,k-1 The CPU occupancy rate of the device running the game. occtop indicates the upper limit of the occupancy rate. bs Indicates the set duration, α1, α2, and α3 represent the weight of similarity, the weight of the number of clicks, and the weight of CPU occupancy respectively; T represents the length of the time interval;

[0064] In calculating t k After that, a countdown is performed, and the countdown duration is the k-th screenshot interval t k .

[0065] The screenshot interval of the present invention is not a fixed value, but is related to the sim k-1,k-2 , nclic and occrat are related to these three parameters. If sim k-1,k-2 The smaller it is, the larger the nclic is, the smaller the occrat is, the smaller the screenshot interval is, and the higher the frequency of screenshots within a given time length; if sim k-1,k-2The larger the value, the smaller the nclic value, and the larger the occrat value, the longer the screenshot interval, and the lower the frequency of screenshots within a given time period. This method of calculating the screenshot interval can comprehensively consider multiple factors to obtain a more reasonable value. Because if a fixed screenshot interval is set, if the screenshot interval is too small, a high frequency of screenshot operations may still be performed when the CPU occupancy rate of the device running the game is high, or when the degree of change in the game screen is small, resulting in the need to occupy more CPU computing resources, affecting the smoothness and realistic effect of the game screen, and causing the overall power consumption of the device to be too high; if the screenshot interval is too large, the screenshot cannot be taken in time, which may cause the user to miss the game strategy in some scenes, affecting the gaming experience. Therefore, the present invention sets an adaptive screenshot interval to achieve a better overall effect.

[0066] Specifically, the upper limit of the second similarity is 1. The second similarity can be calculated using cosine similarity.

[0067] Optional, the default value is 100.

[0068] Specifically, the upper limit of the occupancy rate is 100%.

[0069] Optional, set the duration to 5 seconds.

[0070] Optionally, the weight of similarity, the weight of the number of clicks, and the weight of CPU usage are 0.4, 0.4, and 0.2, respectively.

[0071] Optionally, the time interval is 30 seconds long.

[0072] Preferably, the pre-stored strategy images are screenshot images of the game included in the game strategy.

[0073] Specifically, a game guide generally includes screenshots and text descriptions of the game, so that users who refer to the game guide can understand the content of the game guide more clearly.

[0074] Preferably, calculating the first similarity between the screenshot image and the pre-stored strategy image includes:

[0075] For the screenshot image P1 and the pre-stored strategy image P2, the calculation process of the first similarity between P1 and P2 is as follows:

[0076] If the resolution of P1 is greater than that of P2, P1 is reduced to the same resolution as P2 to obtain image P1b, and the first similarity between P1b and P2 is calculated;

[0077] If the resolution of P1 is smaller than that of P2, P2 is reduced to the same resolution as P1 to obtain image P2b, and a first similarity between P1 and P2b is calculated;

[0078] If the resolution of P1 is the same as the resolution of P2, the first similarity between P1 and P2 is directly calculated.

[0079] Specifically, since the resolutions of P1 and P2 may be different, in order to improve the accuracy of the calculation result of the first similarity, the present invention adjusts the resolutions of the two to be the same before calculating the first similarity.

[0080] Preferably, for two screenshot images Pb and strategy images Pc with the same resolution, the first similarity between Pb and Pc is calculated as follows:

[0081] Calculate Pb and Pc using the preset mask image to obtain images Pb1 and Pc1;

[0082] Calculate the number of sampling areas;

[0083] A first similarity between Pb1 and Pc1 is calculated based on the number of sampling regions.

[0084] Specifically, when there are a large number of game guides, the number of pre-stored guide images is also large. If similarity is calculated based on all pixels, it will obviously be difficult to obtain a similarity result. Therefore, after the device running the game obtains a screenshot image, it can send the screenshot image to a server that stores pre-stored guide images. The server then performs a first similarity calculation based on game guides released within a set time period (e.g., the past year).

[0085] In addition, the present invention also calculates the first similarity by obtaining a sampling area in the image obtained after mask image processing, which can avoid calculating the first similarity for areas containing the same content, and at the same time effectively reduces the number of pixels involved in the calculation process, further improving the calculation efficiency of the similarity.

[0086] Optionally, Pb and Pc are calculated using a preset mask image to obtain images Pb1 and Pc1, including:

[0087] The preset mask image has the same resolution as Pb; in the mask image, the grayscale value of the pixels in the area where the character in the game is controlled is set to 0, the grayscale value of the pixels in the display area of ​​the character in the game is set to 0, and the grayscale value of the remaining pixels is set to 1;

[0088] The process of obtaining Pb1 is as follows:

[0089] Step 1: Perform the following processing on each pixel in Pb:

[0090] For pixel z1 in Pb, obtain pixel z2 in the mask image with the same coordinates as z1, calculate the product G1 of the grayscale value of z1 and the grayscale value of z2, and modify the grayscale value of z1 to G1 in Pb;

[0091] Step 2:

[0092] Delete the pixels with grayscale value 0 in Pb to obtain image Pb1;

[0093] The process of obtaining Pc1 is as follows:

[0094] Step 1: Perform the following processing on each pixel in Pc:

[0095] For pixel z3 in Pc, obtain pixel z4 in the mask image with the same coordinates as z3, calculate the product G2 of the grayscale value of z3 and the grayscale value of z4, and modify the grayscale value of z3 in Pc to G2;

[0096] Step 2:

[0097] Delete the pixels with grayscale value 0 in Pc to obtain image Pc1.

[0098] Specifically, the areas for controlling in-game characters typically appear in the lower left and lower right corners of the game interface. The lower left corner typically contains virtual buttons for controlling character movement, while the lower right corner contains virtual buttons for controlling character skill releases. The in-game character display area is typically located near the center of the bottom edge of the game interface. The corresponding pixels in the mask image can be adaptively adjusted based on the actual game conditions.

[0099] Preferably, calculating the number of sampling areas includes:

[0100] For the screenshot image P obtained after the countdown corresponding to the kth screenshot interval ends k , based on P k The number of sampling areas numsamp during the calculation of the first similarity k The calculation formula is:

[0101]

[0102] numsamp bs represents the upper limit of the number of preset sampling areas, β1 and β2 are and The weight of .

[0103] The number of sampling areas of the present invention can follow the sim k-1,k-2 and t k changes with the changes in sim k-1,k-2 The larger the t k The larger the value, the smaller the degree of change in the game screen. Therefore, the present invention will appropriately increase the number of sampling areas to further improve the accuracy of the calculated first similarity. k-1,k-2 The smaller the t k The smaller the value, the greater the degree of change in the game screen. Therefore, the present invention appropriately reduces the number of sampling areas to achieve higher calculation efficiency for the first similarity. Therefore, the formula for calculating the number of sampling areas in the present invention can adaptively balance calculation efficiency and accuracy, thereby ensuring overall accuracy while improving overall calculation efficiency.

[0104] Optionally, the values ​​of β1 and β2 are both 0.5.

[0105] Optionally, the upper limit of the preset number of sampling areas is 30.

[0106] Preferably, calculating the first similarity between Pb1 and Pc1 based on the number of sampling regions includes:

[0107] Using numsamp Pb Indicates the number of sampling areas corresponding to Pb, and randomly selects numsamp from Pb1 Pb Pixels are stored in the set Pxu1;

[0108] Get the sampling area corresponding to each pixel in Pxu1 respectively;

[0109] The first similarity is calculated using the following formula:

[0110]

[0111] simfir means the first similarity, sim i Indicates the sampling area sampare corresponding to the i-th pixel in Pxu1 i and the area sampare in Pc1 i,Pc1 similarity between

[0112] sampare i,Pc1 For PC1 and sampare i The area formed by all pixels with the same coordinates as the pixels in .

[0113] Optionally, the cosine similarity calculation method or the Jaccard similarity coefficient calculation method can be used to calculate the sampling area samparei and the area sampare in Pc1 i,Pc1 The similarity between them.

[0114] Optionally, obtain the sampling area corresponding to each pixel in Pxu1 separately, including:

[0115] For the pixel point px in Pxu1, the area formed by the pixel points in the square with a side length of R and centered at px is used as the sampling area corresponding to px.

[0116] Optional, the value of R is 9.

[0117] Preferably, obtaining a list of corresponding game strategies according to the pre-stored strategy image corresponding to the largest first similarity includes:

[0118] Each game guide contains at least one guide image;

[0119] A list of game guides is generated based on hyperlink titles of all game guides containing pre-stored guide images corresponding to the largest first similarities.

[0120] Preferably, generating a list of game guides based on the hyperlink titles of all game guides containing pre-stored guide images corresponding to the largest first similarity comprises:

[0121] All hyperlink titles of game guides containing the pre-stored guide image corresponding to the largest first similarity are sorted in descending order according to the release time of the game guides to obtain a list of game guides.

[0122] Second, as Figure 2 ,The present invention provides a game interface data recognition system, including a screenshot module, a recognition module, a judgment module, an acquisition module and a detection module;

[0123] The screenshot module is used to take screenshots of the game interface using an adaptive screenshot interval to obtain screenshot images;

[0124] The recognition module is used to recognize the screenshot image and calculate the first similarity between the screenshot image and the pre-stored strategy image;

[0125] The judging module is used to judge whether the maximum first similarity is greater than a set similarity threshold, and if so, obtain the pre-stored strategy image corresponding to the maximum first similarity;

[0126] The acquisition module is used to acquire a list of corresponding game strategies according to the pre-stored strategy image corresponding to the largest first similarity;

[0127] The detection module is used to close the floating ball displayed on the edge of the game interface after detecting that the user clicks on the floating ball, and display the list of game strategies in the form of a pop-up window; after detecting that the user clicks on the shrink button of the pop-up window, close the pop-up window and redisplay the floating ball on the edge of the game interface.

[0128] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for identifying game interface data, characterized in that: include: S1, taking a screenshot of the game interface using an adaptive screenshot interval to obtain a screenshot image; Use adaptive screenshot intervals to take screenshots of the game interface, including: After the countdown corresponding to the k-1th screenshot interval ends, take a screenshot of the game interface. After the screenshot is completed, calculate the kth screenshot interval , The calculation formula is: represents the second similarity between the screenshot images obtained after the countdown corresponding to the k-1th and k-2th screenshot intervals ends, represents the upper limit of the second similarity, Indicates the preset value. Indicates the time interval The number of times the user clicks on the floating ball, Indicates the moment when the countdown corresponding to the k-1th screenshot interval ends. Indicates The CPU usage of the device running the game, Indicates the upper limit of the occupancy rate. Indicates the duration of the setting. 、 and They represent the weight of similarity, the weight of the number of clicks, and the weight of CPU usage respectively; T represents the length of the time interval; In calculating After that, a countdown is performed, and the countdown duration is the kth screenshot interval ; S2, identifying the screenshot image and calculating a first similarity between the screenshot image and a pre-stored strategy image; S3, determining whether the maximum first similarity is greater than a set similarity threshold, and if so, obtaining a pre-stored strategy image corresponding to the maximum first similarity; S4, obtaining a corresponding game strategy list according to the pre-stored strategy image corresponding to the largest first similarity; S5, after detecting that the user clicks on the floating ball displayed at the edge of the game interface, close the floating ball and display the list of game strategies in the form of a pop-up window; after detecting that the user clicks on the shrink button of the pop-up window, close the pop-up window and display the floating ball again at the edge of the game interface.

2. A method for identifying game interface data according to claim 1, characterized in that: The edges of the game interface include the top edge, bottom edge, left edge and right edge.

3. A method for identifying game interface data according to claim 2, characterized in that: The floating ball displayed on the edge of the game interface includes a floating ball displayed on the upper edge.

4. A method for identifying game interface data according to claim 2, characterized in that: The pre-stored guide images are screenshot images of the games included in the game guide.

5. A method for identifying game interface data according to claim 1, characterized in that: Calculating the first similarity between the screenshot image and the pre-stored strategy image includes: For the screenshot image P1 and the pre-stored strategy image P2, the calculation process of the first similarity between P1 and P2 is as follows: If the resolution of P1 is greater than that of P2, P1 is reduced to the same resolution as P2 to obtain image P1b, and the first similarity between P1b and P2 is calculated; If the resolution of P1 is smaller than that of P2, P2 is reduced to the same resolution as P1 to obtain image P2b, and a first similarity between P1 and P2b is calculated; If the resolution of P1 is the same as the resolution of P2, the first similarity between P1 and P2 is directly calculated.

6. A method for identifying game interface data according to claim 5, characterized in that: For two screenshot images Pb and strategy images Pc with the same resolution, the first similarity between Pb and Pc is calculated as follows: Calculate Pb and Pc using the preset mask image to obtain images Pb1 and Pc1; Calculate the number of sampling areas; A first similarity between Pb1 and Pc1 is calculated based on the number of sampling regions.

7. A method for identifying game interface data according to claim 1, characterized in that: Obtaining a corresponding game strategy list according to the pre-stored strategy image corresponding to the largest first similarity includes: Each game guide contains at least one guide image; A list of game guides is generated based on hyperlink titles of all game guides containing pre-stored guide images corresponding to the largest first similarities.

8. A method for identifying game interface data according to claim 1, characterized in that: A list of game guides is generated based on hyperlink titles of all game guides containing pre-stored guide images corresponding to the largest first similarity, including: All hyperlink titles of game guides containing the pre-stored guide image corresponding to the largest first similarity are sorted in descending order according to the release time of the game guides to obtain a list of game guides.

9. A game interface data recognition system, characterized in that: Including screenshot module, recognition module, judgment module, acquisition module and detection module; The screenshot module is used to take screenshots of the game interface using an adaptive screenshot interval to obtain screenshot images; Use adaptive screenshot intervals to take screenshots of the game interface, including: After the countdown corresponding to the k-1th screenshot interval ends, take a screenshot of the game interface. After the screenshot is completed, calculate the kth screenshot interval , The calculation formula is: represents the second similarity between the screenshot images obtained after the countdown corresponding to the k-1th and k-2th screenshot intervals ends, represents the upper limit of the second similarity, Indicates the preset value. Indicates the time interval The number of times the user clicks on the floating ball, Indicates the moment when the countdown corresponding to the k-1th screenshot interval ends. Indicates The CPU usage of the device running the game, Indicates the upper limit of the occupancy rate. Indicates the duration of the setting. 、 and They represent the weight of similarity, the weight of the number of clicks, and the weight of CPU usage respectively; T represents the length of the time interval; In calculating After that, a countdown is performed, and the countdown duration is the kth screenshot interval ; The recognition module is used to recognize the screenshot image and calculate the first similarity between the screenshot image and the pre-stored strategy image; The judging module is used to judge whether the maximum first similarity is greater than a set similarity threshold, and if so, obtain the pre-stored strategy image corresponding to the maximum first similarity; The acquisition module is used to acquire a list of corresponding game strategies according to the pre-stored strategy image corresponding to the largest first similarity; The detection module is used to close the floating ball displayed on the edge of the game interface after detecting that the user clicks on the floating ball, and display the list of game strategies in the form of a pop-up window; after detecting that the user clicks on the shrink button of the pop-up window, close the pop-up window and redisplay the floating ball on the edge of the game interface.

Images