Jigsaw image data processing method and device

Abstract

The invention provides a jigsaw image data processing method and device. The method comprises the steps of obtaining a target image time consumption type currently corresponding to a target user account of a jigsaw game application; the jigsaw image with the image time type being the target image time type is obtained from all the jigsaw images to serve as a current target jigsaw image, and the image time type of the jigsaw image is determined after image classification is conducted on the jigsaw image based on a content segmentation mode; adding the target jigsaw image into an image recommendation list corresponding to the jigsaw game application and aiming at the target user account; according to the method and the device, the efficiency and the reliability of jigsaw image processing can be effectively improved, the timeliness and the pertinence of jigsaw image pushing can be effectively improved, the personalized requirement of a user for jigsaw completion time preference can be met, and the jigsaw image processing efficiency and the jigsaw image pushing effectiveness can be improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a puzzle image data processing method and device. Background Art

[0002] Jigsaw puzzles are a popular intellectual game that splits a complete image into sub-blocks of the same size to form a puzzle image. Users need to reassemble the jigsaw puzzle sub-blocks in a shuffled order into a complete puzzle image. In order to meet the needs of different users, puzzle images are often classified based on the size of the puzzle sub-blocks and the puzzle content. In order to adapt to mobile devices, puzzle images are often split into sub-blocks of sizes such as 8*8, 10*10, and 12*12. Therefore, it is necessary to process the puzzle images to recommend them to users, thereby meeting the personalized needs of users.

[0003] At present, existing jigsaw image data processing methods usually classify jigsaw images based on their content, for example, into types including architecture, food, animation, and animals. Jigsaw game providers will push them based on users' preferences for image content types.

[0004] However, since different users have preferences for the single completion time of jigsaw puzzle games, and the existing jigsaw puzzle image data processing method can only adapt to the user's type preference, but not the duration preference, it is unable to meet the user's personalized needs in a targeted manner. If the average completion time of historical jigsaw puzzle images is calculated and recommended to the user each time the user's time preference is obtained, it will not only affect the efficiency of jigsaw puzzle image processing, but also fail to push new jigsaw puzzle images, which will affect the efficiency of pushing jigsaw puzzle images to users and reduce user stickiness.

[0005] Therefore, there is an urgent need to design a jigsaw image data processing method that can not only meet the user's personalized needs for jigsaw puzzle completion time preferences, but also improve the jigsaw image processing efficiency and push effectiveness. Summary of the invention

[0006] In view of this, the embodiments of the present application provide a puzzle image data processing method and device to eliminate or improve one or more defects existing in the prior art.

[0007] One aspect of the present application provides a method for processing jigsaw puzzle image data, comprising:

[0008] Get the target image time type currently corresponding to the target user account of the jigsaw puzzle game application;

[0009] Obtaining a puzzle image whose image duration type is the target image duration type from each of the puzzle images as the current target puzzle image, wherein the image duration type of the puzzle image is determined after classifying the puzzle image based on content segmentation;

[0010] The target jigsaw puzzle image is added to an image recommendation list corresponding to the jigsaw puzzle game application and for the target user account.

[0011] In some embodiments of the present application, the step of obtaining the target image time type currently corresponding to the target user account of the jigsaw puzzle game application includes:

[0012] Obtaining an average jigsaw puzzle game time corresponding to a target user account of the jigsaw puzzle game application;

[0013] From the correspondence between each preset image time type and each preset time range, the image time type corresponding to the preset time range where the average time of the jigsaw puzzle game is located is determined, and the image time type has been determined as the target image time type currently corresponding to the target user account.

[0014] In some embodiments of the present application, before obtaining the target image time type currently corresponding to the target user account of the jigsaw game application, the method further includes:

[0015] Preprocessing the puzzle image to delete each puzzle sub-block in the boundary area of ​​the puzzle image to obtain a corresponding initialization image;

[0016] Performing image segmentation on the initialization image, and performing partition marking on each segmented area of ​​the initialization image obtained by the image segmentation;

[0017] If there are different segmented areas belonging to the same content object among the segmented areas, the different segmented areas belonging to the same content object are merged to form a merged area, and a unique identifier is re-set for the merged area;

[0018] The regions with the same current identifier are taken as a set, and the average feature vectors corresponding to each of the sets are extracted. If the similarity between the average feature vectors corresponding to different sets meets the similarity threshold, the sets with different similarities meeting the similarity threshold are merged into one set, and the unique identifier is reset for the merged set;

[0019] Extracting features of the puzzle image according to each of the current sets to obtain feature data of the puzzle image;

[0020] The image time type to which the puzzle image belongs is determined according to the feature data of the puzzle image.

[0021] In some embodiments of the present application, the performing image segmentation on the initialization image and performing partition marking on each segmented area of ​​the initialization image obtained by the image segmentation includes:

[0022] Input the initialization image into a preset image segmentation model so that the image segmentation model outputs corresponding segmentation areas corresponding to the initialization image; wherein the image segmentation model includes: a SAM model;

[0023] Each of the partitioned areas is marked as a partition, so that each of the partitioned areas corresponds to a different unique identifier.

[0024] In some embodiments of the present application, the regions with the same current identifier are taken as a set, and the average feature vectors corresponding to each of the sets are extracted. If the similarity between the average feature vectors corresponding to different sets meets the similarity threshold, the sets with different similarities meeting the similarity threshold are merged into one set, and the unique identifier is re-set for the merged set, including:

[0025] The regions with the same current identifier are taken as a set, and each of the image sub-blocks in each of the sets is sequentially input into a preset image feature extraction model, so that the image feature extraction model outputs a multi-dimensional feature vector corresponding to each of the image sub-blocks; wherein the image feature extraction model includes: a Resnet model and / or a VGG model;

[0026] According to the multidimensional feature vectors corresponding to the image sub-blocks in the sets, respectively, determining the average feature vectors corresponding to the sets;

[0027] Determine whether the similarity between the average feature vectors corresponding to different sets satisfies the similarity threshold; if so, merge the sets whose similarities satisfy the similarity threshold into one set, and reset the unique identifier for the merged set so that each image sub-block in the same set is marked with the same unique identifier.

[0028] In some embodiments of the present application, extracting features from the puzzle image according to each of the current sets to obtain feature data of the puzzle image includes:

[0029] Obtaining the number of the current sets of the puzzle image and the number of image sub-blocks corresponding to each set, wherein the number of image sub-blocks corresponding to the set is the area of ​​the set;

[0030] Sorting the sets in descending order of area, and determining the mean square error of the areas of the first three sets after sorting; and determining the mean square error of the areas of all the sets of the puzzle image at present;

[0031] According to the number of the sets and the mean square error of the areas of the first three sets, the proportion of the total area of ​​the first three sets is determined; and the number of the sets whose areas are less than the area threshold is determined, so as to obtain characteristic data including the number of the sets, the areas corresponding to each of the sets, the mean square error of the areas of the first three sets, the mean square error of the areas of all the sets, the proportion of the total area of ​​the first three sets and the number of the sets whose areas are less than the first area threshold.

[0032] In some embodiments of the present application, determining the image time type to which the puzzle image belongs according to the feature data of the puzzle image includes:

[0033] Determine whether the area of ​​the largest set among the sets corresponding to the puzzle image is equal to or greater than a second area threshold, and if so, determine that the image time type to which the puzzle image belongs is: the first type; wherein the second area threshold is greater than the first area threshold;

[0034] If the area of ​​the largest set among the sets corresponding to the puzzle image is smaller than the second area threshold, determining whether the total area ratio of the first three sets corresponding to the puzzle image is equal to or greater than the ratio threshold;

[0035] If the total area ratio of the first three sets corresponding to the puzzle image is equal to or greater than the ratio threshold, then determine whether the feature data corresponding to the puzzle image meets a preset condition, wherein the preset condition includes: the number of sets corresponding to the puzzle image is less than a first number threshold, the area of ​​the set with the largest area is less than a third area threshold, and the mean square deviation of the areas of all the sets is less than a mean square deviation threshold;

[0036] If the feature data corresponding to the puzzle image does not meet the preset condition, the image time type to which the puzzle image belongs is determined to be the second type; if the feature data corresponding to the puzzle image meets the preset condition, the image time type to which the puzzle image belongs is determined to be the third type;

[0037] If the total area ratio of the first three sets corresponding to the puzzle image is less than the ratio threshold, then determining whether the number of sets corresponding to the puzzle image whose area is less than the first area threshold exceeds a second number threshold, wherein the second number threshold is greater than the first number threshold; if so, determining that the image time type to which the puzzle image belongs is: the fourth type;

[0038] Among them, the preset time ranges corresponding to the first type, the second type, the third type to the fourth type respectively decrease in sequence.

[0039] Another aspect of the present application provides a jigsaw puzzle image data processing device, comprising:

[0040] An image time type acquisition module is used to acquire the target image time type currently corresponding to the target user account of the jigsaw game application;

[0041] a puzzle image acquisition module, used for acquiring a puzzle image whose image duration type is the target image duration type from each puzzle image, as the current target puzzle image, wherein the image duration type of the puzzle image is determined after classifying the puzzle image based on content segmentation;

[0042] The puzzle image recommendation module is used to add the target puzzle image to the image recommendation list for the target user account corresponding to the puzzle game application.

[0043] A third aspect of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and a puzzle image data processing method implemented when the processor executes the computer program.

[0044] A fourth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which implements a puzzle image data processing method when the computer program is executed by a processor.

[0045] A fifth aspect of the present application provides a computer program product, including a computer program, which implements the puzzle image data processing method when executed by a processor.

[0046] The jigsaw puzzle image data processing method provided by the present application obtains the target image time type currently corresponding to the target user account of the jigsaw puzzle game application; obtains a jigsaw puzzle image with an image time type of the target image time type from each of the jigsaw puzzle images to serve as the current target jigsaw puzzle image, wherein the image time type of the jigsaw puzzle image is determined after classifying the jigsaw puzzle image based on content segmentation; and adds the target jigsaw puzzle image to the image recommendation list corresponding to the jigsaw puzzle game application for the target user account, which can effectively improve the efficiency and reliability of jigsaw puzzle image processing, and can effectively improve the timeliness and pertinence of jigsaw puzzle image push, which can not only meet the user's personalized needs for jigsaw puzzle completion time preference, but also improve the jigsaw puzzle image processing efficiency and push effectiveness.

[0047] Additional advantages, purposes, and features of the present application will be partially described in the following description, and will become partially apparent to those skilled in the art after studying the following, or may be learned from the practice of the present application. The purposes and other advantages of the present application can be achieved and obtained by the structures specifically pointed out in the specification and the drawings.

[0048] Those skilled in the art will understand that the purposes and advantages that can be achieved by the present application are not limited to the above specifics, and the above and other purposes that can be achieved by the present application will be more clearly understood based on the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings described herein are used to provide a further understanding of the present application, constitute a part of the present application, and do not constitute a limitation of the present application. The components in the drawings are not drawn to scale, but are only for the purpose of illustrating the principles of the present application. In order to facilitate the illustration and description of some parts of the present application, the corresponding parts in the drawings may be enlarged, that is, they may become larger relative to other components in the exemplary device actually manufactured according to the present application. In the drawings:

[0050] Figure 1 4 is a schematic diagram of a first flow chart of a puzzle image data processing method in an embodiment of the present application.

[0051] Figure 2 2 is a schematic diagram of a second flow chart of a puzzle image data processing method in an embodiment of the present application.

[0052] Figure 3 2 is a flow chart of step 020 in the puzzle image data processing method in one embodiment of the present application.

[0053] Figure 4 4 is a flowchart of step 040 in the puzzle image data processing method in one embodiment of the present application.

[0054] Figure 5 4 is a flow chart of step 050 in the puzzle image data processing method in one embodiment of the present application.

[0055] Figure 6 Schematic diagram of the flow of step 060 in the puzzle image data processing method in one embodiment of the present application.

[0056] Figure 7 Schematic diagram of the structure of a jigsaw puzzle image data processing device in one embodiment of the present application.

[0057] Figure 8 This is a flowchart of the process of image classification of puzzle images based on content segmentation in the application example of this application.

[0058] FIG9( a ) is the original image in the application example of this application.

[0059] FIG. 9( b ) is an image segmentation result and puzzle sub-block numbering corresponding to the original image of FIG. 9( a ) in the application example of the present application.

[0060] Fig.10 This is a schematic diagram of a collection of temporarily stacked jigsaw puzzle pieces during a jigsaw puzzle game in an application example of the present application.

[0061] FIG. 11( a ) is a schematic diagram of the segmentation result of an animal target in an application example of the present application.

[0062] FIG11( b ) is a schematic diagram showing the merging results of the target area puzzle sub-block numbers corresponding to the animal target in the application example of the present application.

[0063] FIG12( a ) is the original image of a butterfly in the application example of the present application.

[0064] FIG. 12( b ) is a schematic diagram of the result before the numbers corresponding to the original butterfly image of FIG. 12( a ) are merged.

[0065] FIG. 12( c ) is a schematic diagram showing the result of merging the numbers corresponding to the original butterfly image in FIG. 12( a ).

[0066] Fig.13 It is a schematic diagram of a pile formed after feature extraction of FIG. 11( b ) provided in an application example of the present application.

[0067] FIG. 14( a ) is a first example of a large monomer type puzzle image provided in the application example of the present application.

[0068] FIG14( b ) is a schematic diagram of the characteristic data corresponding to FIG14( a ).

[0069] FIG. 14( c ) is a second example of a large monomer type puzzle image provided in the application example of the present application.

[0070] FIG14( d ) is a schematic diagram of the characteristic data corresponding to FIG14( c ).

[0071] FIG. 15( a ) is a first example of a non-uniform Top3 type puzzle image provided in the application example of the present application.

[0072] FIG15( b ) is a schematic diagram of the characteristic data corresponding to FIG15( a ).

[0073] FIG. 15( c ) is a second example of a non-uniform Top3 type puzzle image provided in the application example of the present application.

[0074] FIG15( d ) is a schematic diagram of the characteristic data corresponding to FIG15( c ).

[0075] FIG. 16( a ) is a first example of a uniformly distributed puzzle image provided in the application example of the present application.

[0076] FIG16( b ) is a schematic diagram of the characteristic data corresponding to FIG16( a ).

[0077] FIG. 16( c ) is a second example of a uniformly distributed puzzle image provided in the application example of the present application.

[0078] FIG16( d ) is a schematic diagram of the characteristic data corresponding to FIG16( c ).

[0079] FIG. 17( a ) is a first example of a puzzle image with multiple small blocks distributed according to an application example of the present application.

[0080] FIG17( b ) is a schematic diagram of the characteristic data corresponding to FIG17( a ).

[0081] FIG. 17( c ) is a second example of a puzzle image with multiple small blocks distributed according to the application example of the present application.

[0082] FIG17( d ) is a schematic diagram of the characteristic data corresponding to FIG17( c ). DETAILED DESCRIPTION

[0083] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the implementation modes and the accompanying drawings. Here, the illustrative implementation modes and descriptions of the present application are used to explain the present application, but are not intended to limit the present application.

[0084] It should also be noted here that in order to avoid obscuring the present application due to unnecessary details, only the structures and / or processing steps closely related to the scheme according to the present application are shown in the accompanying drawings, while other details that are not very relevant to the present application are omitted.

[0085] It should be emphasized that the term “include / comprises” when used herein refers to the presence of features, elements, steps or components, but does not exclude the presence or addition of one or more other features, elements, steps or components.

[0086] It should also be noted that, unless otherwise specified, the term “connection” herein may refer not only to a direct connection but also to an indirect connection involving an intermediate.

[0087] Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. In the accompanying drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

[0088] In order to solve the problem that the existing puzzle image data processing methods cannot simultaneously achieve targeted satisfaction of users' personalized needs, ensure the efficiency of puzzle image processing, and realize the push of new puzzle images, the embodiments of the present application respectively provide a puzzle image data processing method, a puzzle image data processing device for executing the puzzle image data processing method, an electronic device, a computer-readable storage medium, and a computer program product, which can not only meet the users' personalized needs for puzzle completion time preferences, but also improve the puzzle image processing efficiency and push effectiveness.

[0089] The details are described in detail through the following examples.

[0090] Based on this, the embodiment of the present application provides a puzzle image data processing method that can be implemented by a puzzle image data processing device, see Figure 1 , the jigsaw image data processing method specifically includes the following contents:

[0091] Step 100: Obtain the target image time type currently corresponding to the target user account of the jigsaw game application.

[0092] In one implementation of step 100, a user may select an average jigsaw puzzle time of his or her preference in a jigsaw puzzle application, and then send the average jigsaw puzzle time and the user account information corresponding to the user to a jigsaw image data processing device. The jigsaw image data processing device receives the average jigsaw puzzle time and the user account information corresponding to the user to determine a target image time type currently corresponding to a target user account of the jigsaw puzzle application.

[0093] In one or more embodiments of the present application, the user account refers to a unique identifier corresponding to the user in the jigsaw game application, and the same user account corresponds to only one user.

[0094] It can be understood that the average time taken for a jigsaw puzzle game refers to the average time taken by a user to complete a single jigsaw puzzle image.

[0095] Step 200: Obtain a puzzle image whose image duration type is the target image duration type from each of the puzzle images as the current target puzzle image, wherein the image duration type of the puzzle image is determined after classifying the puzzle image based on content segmentation.

[0096] In step 200, a local image library of the puzzle image data processing device pre-stores a plurality of puzzle images corresponding to various image usage types. These puzzle images may preferably be new puzzle images that have not been selected by the user.

[0097] Correspondingly, when a puzzle image among the locally stored puzzle images is selected as the current target puzzle image, the puzzle image and its corresponding image time type can be deleted locally in the image library, so that the target puzzle image pushed to the user each time is an image that the user has not yet adopted.

[0098] Furthermore, in order to prevent the jigsaw image from being consumed quickly, when there is a jigsaw image selected as the current target jigsaw image among the jigsaw images stored locally, the corresponding relationship between the jigsaw image, the image time type corresponding to the jigsaw image, and the target user account can also be constructed locally to mark locally that these jigsaw images have been pushed to the target user account. Based on this, in a preferred method of step 200, it is possible to first find out whether the corresponding relationship between the target user account, the jigsaw image, and the image time type corresponding to the jigsaw image is stored locally. If so, the jigsaw images and their image time types corresponding to the target user account stored locally are first filtered out in the image library, and then the jigsaw images with the image time type of the target image are obtained in the filtered image library as the current target jigsaw image.

[0099] Step 300: Add the target puzzle image to the image recommendation list corresponding to the target user account of the puzzle game application.

[0100] From the above description, it can be seen that the puzzle image data processing method provided in the embodiment of the present application can effectively improve the efficiency and reliability of puzzle image processing, and can effectively improve the timeliness and pertinence of puzzle image push, which can not only meet the user's personalized needs for puzzle completion time preferences, but also improve the puzzle image processing efficiency and push effectiveness.

[0101] In order to further obtain the validity and intelligence of the target image time type currently corresponding to the target user account of the jigsaw puzzle game application, in a jigsaw puzzle image data processing method provided in an embodiment of the present application, see Figure 2 , step 100 in the jigsaw image data processing method specifically includes the following contents:

[0102] Step 110: Obtain the average jigsaw puzzle game time corresponding to the target user account of the jigsaw puzzle application.

[0103] Step 120: Determine the image time type corresponding to the preset time range in which the average time of the jigsaw puzzle game is located from the correspondence between each preset image time type and each preset time range, and determine the image time type as the target image time type currently corresponding to the target user account.

[0104] In order to further improve the effectiveness and reliability of image classification of puzzle images based on content segmentation, so as to further improve the efficiency and reliability of puzzle image processing and meet the personalized needs of users for puzzle completion time preference, in a puzzle image data processing method provided in an embodiment of the present application, see Figure 2 , the jigsaw image data processing method further specifically includes the following contents before step 100:

[0105] Step 010: pre-processing the puzzle image to delete each puzzle sub-block in the boundary area of ​​the puzzle image to obtain a corresponding initialization image.

[0106] It is understandable that in step 010, in the jigsaw puzzle game, the jigsaw sub-blocks located at the edges of the jigsaw puzzle image have distinct characteristics and are easiest to be spliced ​​first. Therefore, in the subsequent calculation, the characteristics of the sub-blocks at the outermost edge are not included in the statistical scope.

[0107] Step 020: performing image segmentation on the initialization image, and performing partition marking on each segmented area of ​​the initialization image obtained by the image segmentation.

[0108] Step 030: If there are different segmented areas belonging to the same content object in each of the segmented areas, the different segmented areas belonging to the same content object are merged to form a merged area, and a unique identifier is re-set for the merged area.

[0109] It is understandable that in step 030, when playing the jigsaw puzzle game, the user will often temporarily put together the jigsaw puzzle sub-blocks that belong to the same target based on visual features such as texture and color. In particular, when there are animals in the jigsaw puzzle image, the user will temporarily pile up the jigsaw puzzle sub-blocks that belong to the same animal to facilitate subsequent splicing. In order to simulate this user experience, it is necessary to first segment the content objects such as animals in the image, and merge the numbers of the jigsaw puzzle sub-blocks where the content objects such as animals are located and change them to the same number. If a jigsaw puzzle sub-block contains content objects such as animals, regardless of the size of the area, the number of the jigsaw puzzle sub-block is uniformly changed to the number of the jigsaw puzzle sub-block where the animal is located.

[0110] Step 040: The areas with the same current identifier are taken as a set, and the average feature vectors corresponding to each of the sets are extracted. If the similarity between the average feature vectors corresponding to different sets meets the similarity threshold, the sets with different similarities that meet the similarity threshold are merged into one set, and the unique identifier of the merged set is reset.

[0111] Step 050: extracting features of the puzzle image according to each of the current sets to obtain feature data of the puzzle image.

[0112] Step 060: Determine the image time type to which the puzzle image belongs based on the feature data of the puzzle image.

[0113] In order to further improve the reliability and effectiveness of performing image segmentation on the initialization image and performing partition marking on each segmented area of ​​the initialization image obtained by the image segmentation, in a jigsaw image data processing method provided in an embodiment of the present application, see Figure 3 , step 020 in the jigsaw image data processing method specifically includes the following contents:

[0114] Step 021: Input the initialization image into a preset image segmentation model so that the image segmentation model outputs each segmentation area corresponding to the initialization image; wherein the image segmentation model includes: a SAM model.

[0115] It is understandable that the SAM model (Segment Anything Model) is a large image segmentation model open sourced by Meta that can segment any object in any image without any annotation. It uses a web browser interface to predict masks using user-provided cues (such as punctuation, boxes, and text) to segment objects in the image.

[0116] Step 022: Partition and mark each of the partitioned areas so that each of the partitioned areas corresponds to a different unique identifier.

[0117] Specifically, in the actual jigsaw puzzle process, since the positions of the jigsaw puzzle sub-blocks are disrupted, users often select the target sub-blocks based on the integrity of the target in the image and the color, texture and other characteristics of different targets. To this end, the SAM (Segment Anything Model) model is used to segment the image, distinguish different targets in the image, and assign the same number to the jigsaw puzzle sub-blocks belonging to the same segmented area. If a jigsaw puzzle sub-block contains multiple different segmented areas, it is represented by the segmented area number with the largest area. The number size has no special meaning, but jigsaw puzzle sub-blocks with the same number are considered to be a collection that is easily temporarily stacked together during the jigsaw puzzle game. Temporarily stacking similar jigsaw puzzle sub-blocks together is a common strategy in jigsaw puzzle games.

[0118] In order to further improve the reliability and effectiveness of the image sub-block number merging, in a jigsaw image data processing method provided in an embodiment of the present application, see Figure 4 , step 040 in the jigsaw image data processing method specifically includes the following contents:

[0119] Step 041: taking the regions with the same current identifier as a set, and inputting the image sub-blocks in each of the sets into a preset image feature extraction model in sequence, so that the image feature extraction model outputs a multi-dimensional feature vector corresponding to each of the image sub-blocks; wherein the image feature extraction model includes: a Resnet model and / or a VGG model;

[0120] Step 042: According to the multi-dimensional feature vectors corresponding to the image sub-blocks in each of the sets, respectively determine the average feature vectors corresponding to the sets.

[0121] Step 043: Determine whether there are different sets whose corresponding average feature vectors have similarities that satisfy a similarity threshold. If so, merge the sets whose similarities satisfy different similarity thresholds into one set, and reset the unique identifier for the merged set so that each of the image sub-blocks in the same set is marked with the same unique identifier. If not, do not merge the sets.

[0122] It is understandable that the ResNet model (deep residual network) is a convolutional neural network that solves the gradient vanishing and gradient exploding problems in deep neural network training by introducing residual learning, allowing the network to learn features and representations more deeply and accurately. The VGG (Visual Geometry Group model) model usually contains multiple convolutional layers and fully connected layers, and it can perform well in a variety of image recognition and classification tasks, especially when using pre-trained models for transfer learning.

[0123] That is to say, during the jigsaw puzzle game, users often temporarily stack image sub-blocks with similar colors and textures together. This application names a set of image sub-blocks with the same number as a pile. In order to simulate the experience during this game, the features of each image sub-block are first calculated separately. In practical applications, the features of each image sub-block can be extracted through neural networks such as Resnet or VGG, and represented as a multi-dimensional vector. Then, the average feature vector corresponding to each pile is calculated. Finally, the cosine similarity between the average feature vectors of each two piles is calculated. If the similarity is greater than the specified threshold, the numbers of the two piles are merged.

[0124] In order to further improve the reliability and effectiveness of feature extraction of the puzzle image, in a puzzle image data processing method provided in an embodiment of the present application, see Figure 5 , step 050 in the jigsaw image data processing method specifically includes the following contents:

[0125] Step 051: obtaining the number of the current sets of the puzzle image and the number of image sub-blocks corresponding to each set, wherein the number of image sub-blocks corresponding to the set is the area of ​​the set;

[0126] Step 052: sorting the sets in descending order of area, and determining the mean square error of the areas of the first three sets after sorting; and determining the mean square error of the areas of all the sets of the puzzle image at present;

[0127] Step 053: Determine the proportion of the total area of ​​the first three sets according to the number of the sets and the mean square error of the areas of the first three sets; and determine the number of the sets whose areas are less than the area threshold, so as to obtain characteristic data including the number of the sets, the areas corresponding to each of the sets, the mean square error of the areas of the first three sets, the mean square error of the areas of all the sets, the proportion of the total area of ​​the first three sets and the number of the sets whose areas are less than the first area threshold.

[0128] In order to further improve the reliability and effectiveness of determining the image time type to which the puzzle image belongs based on the feature data of the puzzle image, in a puzzle image data processing method provided in an embodiment of the present application, see Figure 6 , step 060 in the jigsaw image data processing method specifically includes the following contents:

[0129] Step 061: Determine whether the area of ​​the largest set among the sets corresponding to the puzzle image is equal to or greater than the second area threshold. If so, execute step 062; if not (i.e., if the area of ​​the largest set among the sets corresponding to the puzzle image is less than the second area threshold), execute step 063.

[0130] Step 062: Determine that the image time type to which the puzzle image belongs is: the first type; wherein the second area threshold is greater than the first area threshold;

[0131] Step 063: Determine whether the total area ratio of the first three sets corresponding to the puzzle image is equal to or greater than the ratio threshold; if so (if the total area ratio of the first three sets corresponding to the puzzle image is equal to or greater than the ratio threshold), execute step 064; if not (if the total area ratio of the first three sets corresponding to the puzzle image is less than the ratio threshold), execute step 067.

[0132] Step 064: Determine whether the feature data corresponding to the puzzle image meets the preset conditions. If not (if the feature data corresponding to the puzzle image does not meet the preset conditions), execute step 065; if yes (if the feature data corresponding to the puzzle image meets the preset conditions), execute step 066.

[0133] The preset conditions include: the number of the sets corresponding to the puzzle image is less than a first quantity threshold, the area of ​​the set with the largest area is less than a third area threshold, and the mean square error of the areas of all the sets is less than a mean square error threshold.

[0134] Step 065: Determine that the image time type to which the puzzle image belongs is the second type.

[0135] Step 066: Determine that the image time type to which the puzzle image belongs is the third type.

[0136] Step 067: Determine whether the number of the sets corresponding to the puzzle images whose areas are smaller than the first area threshold exceeds the second number threshold. If so, execute step 068; if not, determine the puzzle image as a useless type puzzle image, and store all useless type puzzle images independently.

[0137] Wherein, the second quantity threshold is greater than the first quantity threshold;

[0138] Step 068: Determine that the image time type to which the puzzle image belongs is the fourth type.

[0139] In one or more embodiments of the present application, the preset time ranges corresponding to the first type, the second type, the third type to the fourth type respectively decrease in sequence.

[0140] Among them, the first type can also be called: macromonomer type;

[0141] The second type can also be called: non-uniform Top3 type;

[0142] The third type can also be called: uniform Top3 type or uniform distribution type;

[0143] The fourth type can also be called: multi-small block distribution type.

[0144] In one or more embodiments of the present application, the first area threshold can be set to between 5% and 10% of the total area of ​​the complete puzzle image, preferably 6.25%. For example, for an 8*8 puzzle image, the first area threshold is preferably 4. It can be understood that the area mentioned in the present application refers to the number of segments, such as 8*8 segments, the area is 64. The specific value can be set according to the actual application scenario. The second area threshold can be set to between 40% and 60% of the total area of ​​the complete puzzle image, preferably 54.69%. For example, for an 8*8 puzzle image, the second area threshold is preferably 35.

[0145] The third area threshold may be set to be between 20% and 30% of the total area of ​​the complete puzzle image, preferably 31.25%. For example, for an 8*8 puzzle image, the third area threshold is preferably 20.

[0146] The percentage threshold may be set between 80% and 100%, preferably 90%.

[0147] The first quantity threshold may be set between 20% and 25% of the number of all sets in the complete puzzle image, preferably 23.44%. For example, for an 8*8 puzzle image, the second quantity threshold is preferably 15.

[0148] The second quantity threshold may be set between 25% and 30% of the number of all sets in the complete puzzle image, preferably 26.56%. For example, for an 8*8 puzzle image, the second quantity threshold is preferably 17.

[0149] The mean square error threshold can be set according to the needs. For example, for an 8*8 puzzle image, the mean square error threshold is preferably 5.

[0150] From the software level, the present application also provides a jigsaw puzzle image data processing device for executing all or part of the jigsaw puzzle image data processing method, see Figure 7 The jigsaw image data processing device specifically includes the following contents:

[0151] An image time type acquisition module 10 is used to acquire a target image time type currently corresponding to a target user account of a jigsaw puzzle application;

[0152] A puzzle image acquisition module 20 is used to acquire a puzzle image whose image duration type is the target image duration type from each puzzle image to serve as the current target puzzle image, wherein the image duration type of the puzzle image is determined after classifying the puzzle image based on content segmentation;

[0153] The puzzle image recommendation module 30 is used to add the target puzzle image to the image recommendation list for the target user account corresponding to the puzzle game application.

[0154] The embodiment of the puzzle image data processing device provided in the present application can be specifically used to execute the processing flow of the embodiment of the puzzle image data processing method in the above embodiment. Its functions will not be repeated here, and reference can be made to the detailed description of the above puzzle image data processing method embodiment.

[0155] The part of the jigsaw image data processing device that processes the jigsaw image data can be completed in the server or the client device. The specific selection can be made according to the processing capability of the client device and the limitations of the user's usage scenario. This application does not limit this. If all operations are completed in the client device, the client device may also include a processor for specific processing of the jigsaw image data processing.

[0156] The client device mentioned above may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission with the server. The server may include a server on the task scheduling center side, and other implementation scenarios may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, or a server cluster consisting of multiple servers, or a server structure of a distributed device.

[0157] The server and the client device may communicate with each other using any suitable network protocol, including network protocols that have not yet been developed on the date of filing this application. Network protocols may include, for example, TCP / IP, UDP / IP, HTTP, HTTPS, etc. Of course, network protocols may also include, for example, RPC (Remote Procedure Call Protocol) and REST (Representational State Transfer) protocols used on top of the above protocols.

[0158] From the above description, it can be seen that the puzzle image data processing device provided in the embodiment of the present application can effectively improve the efficiency and reliability of puzzle image processing, and can effectively improve the timeliness and pertinence of puzzle image push, which can not only meet the user's personalized needs for puzzle completion time preferences, but also improve the puzzle image processing efficiency and push effectiveness.

[0159] In order to further illustrate the embodiment of the above-mentioned puzzle image data processing method, the present application also provides a specific application example of image classification based on content segmentation of puzzle images corresponding to steps 010 to 060 in the puzzle image data processing method, see Figure 8 , the image classification process of puzzle images based on content segmentation specifically includes the following contents:

[0160] Based on the image content, it is divided into common types including architecture, food, animation, animals, etc., and the game manufacturer will make recommendations based on the user's preference for image content. In addition to this routine operation, in order to maintain the competitiveness of the product, game manufacturers also need to identify the preferences of different users based on a more detailed user experience, classify and stratify users, and recommend different puzzle levels for users of different levels and preferences. To this end, the application example of this application can conduct a more detailed quantitative analysis of the user's time preference for jigsaw games, classify the experience of jigsaw games with more fine-grained features, achieve more accurate user classification and content push, and ultimately increase the number of daily active users and user stickiness of the game.

[0161] Based on this, in order to solve the problem of insufficient precision in user classification and content push of jigsaw puzzle games, the application example of this application provides a method for image classification of jigsaw puzzle images based on image content understanding, segmentation and statistics:

[0162] 1. Preprocessing (corresponding to the above step 010)

[0163] In the jigsaw puzzle game, the jigsaw puzzle pieces located on the edges of the image have distinct characteristics and are easiest to be assembled first. Therefore, in the later calculations, the features of the pieces on the outermost edge are not included in the statistical scope.

[0164] In one example, the original image of the puzzle image may be an image of 8*8 or 10*10 size.

[0165] 2. Image segmentation and numbering (corresponding to the above step 020)

[0166] In the actual jigsaw puzzle process, since the positions of the jigsaw puzzle pieces are disrupted, users often select target sub-blocks based on the integrity of the target in the image and the color, texture and other characteristics of different targets. To this end, the SAM (SegmentAnything Model) model is used to segment the image, distinguish different targets in the image, and assign the same number to the jigsaw puzzle pieces belonging to the same segmented area, as shown in Figure 9(a) and Figure 9(b). If a jigsaw puzzle piece contains multiple different segmented areas, it is represented by the segmented area number with the largest area. The number size has no special meaning, but jigsaw puzzle pieces with the same number are considered to be a collection that is easily temporarily stacked together during the jigsaw puzzle game. Temporarily stacking similar puzzle pieces together is a common strategy in jigsaw puzzle games. For example Fig.10 As shown in the two red frame areas, two sets of sub-blocks are temporarily stacked together.

[0167] 3. Single target number merging (corresponding to the above step 030)

[0168] When playing jigsaw puzzles, users often temporarily put together puzzle pieces that belong to the same target based on visual features such as texture and color. In particular, when there are animals in the image, users will temporarily pile up puzzle pieces that belong to the same animal to facilitate subsequent splicing. In order to simulate this user experience, it is necessary to first split the animal target (i.e., a type of content object) in the image and merge the puzzle piece numbers where the animal target is located to the same number, as shown in Figure 11(a) and Figure 11(b).

[0169] Blocks 140 and 131 in Figure 11(a) are both numbered 19 in Figure 11(b), indicating that the sub-block numbers of the animal target segmentation area are unified. If a puzzle sub-block contains an animal target area, regardless of the size of the area, the number of the puzzle sub-block is uniformly changed to the puzzle sub-block number where the animal target is located. As shown in Figure 11(a), the sub-block in the second row and first column is numbered 80, which is different from the number 140 of the puppy head area. However, in Figure 11(b), the same sub-block in the second row and first column, numbered 19, is unified with the number of the puppy area.

[0170] 4. Image sub-block number merging (corresponding to the aforementioned step 040)

[0171] During the jigsaw puzzle game, users often temporarily stack image sub-blocks with similar colors and textures together. The set of image sub-blocks with the same number is named a pile. In order to simulate the experience in this game process, the features of each image sub-block are first calculated separately. In practical applications, the features of each image sub-block can be extracted through neural networks such as Resnet or VGG and represented as a multi-dimensional vector. Then, the average feature vector corresponding to each pile is calculated. Finally, the cosine similarity between the average feature vectors of each two piles is calculated. If the similarity is greater than the specified threshold, the numbers of the two piles are merged. As shown in the original butterfly image of Figure 12(a), Figure 12(b) is the result before the numbers corresponding to the original butterfly image of Figure 12(a) are merged, and Figure 12(c) is the result after the numbers corresponding to the original butterfly image of Figure 12(a) are merged. As can be seen from Figure 12(c), after the merger, the numbers of the upper and lower butterfly areas are the same.

[0172] 5. Feature extraction (corresponding to the above step 050)

[0173] In one or more embodiments of the present application, a collection may be referred to as a heap.

[0174] Correspondingly, after the above steps, the entire image is divided into several piles. The feature data to be extracted next include: the number of piles, the area of ​​each pile (i.e. the total number of image sub-blocks contained), the mean square error of the areas of the top three piles, the mean square error of the areas of all piles, the total area ratio of the top three piles (i.e. the size of the top three piles divided by the total number of image sub-blocks), the number of piles with an area less than 4, etc. Fig.13 As shown, there are three piles numbered 19, 24, and 20. The area of ​​pile numbered 24 is 3, the area of ​​pile numbered 20 is 13, and the area of ​​pile numbered 19 is 48. In this example, the total area of ​​the top three piles accounts for 100%.

[0175] 6. Quantitative classification (corresponding to the above step 060)

[0176] Based on the features extracted above, taking the 8*8 jigsaw puzzle game as an example, the game experience is divided into the following typical types:

[0177] 1) Large Monomer Type: The area of ​​the first largest pile is more than 35. That is, the pile ranked first in area has more than 35 puzzle pieces. Figures 14(a) to 14(d) As shown, this type of large monomer type may generally take users a long time, and users will often complete the splicing of large monomer areas in the second half of the game.

[0178] 2) Top3 type: The sum of the areas of the top three heaps accounts for more than 90%, and the area of ​​the first heap is less than 35. According to the mean square error of the number of image sub-blocks in the top three heaps, this type can be further divided into uniform Top3 and non-uniform Top3 types. Figures 15(a) to 15(d) As shown, the non-uniform Top3 type puzzle takes longer time than the uniform Top3 type.

[0179] 3) Uniform distribution: The number of piles is less than 15, the area of ​​the largest pile is less than 20, and the mean square error of the areas of all piles is less than 5. Since the area distribution of the piles is relatively uniform and the area difference between the piles is not large, such as Figures 16(a) to 16(d) As shown, this type of puzzle is relatively simple and takes less time.

[0180] 4) Multiple small-block distribution type: There are more than 17 piles with an area less than 4. Figures 17(a) to 17(d) As shown in Figure 1, this type of puzzle contains many small pieces with independent textures and colors. If the piles in the puzzle are all different colors, the game becomes easy. However, if there are many piles of similar colors and textures in the puzzle, the game becomes difficult.

[0181] The present application also provides an electronic device, which may include a processor, a memory, a receiver and a transmitter. The processor is used to execute the jigsaw image data processing method mentioned in the above embodiment, wherein the processor and the memory may be connected via a bus or other means, such as a bus connection. The receiver may be connected to the processor and the memory via a wired or wireless manner. It is understood that Figure 9(a) to Figure 17(b) The numbers in the figure are only examples and can be set according to the actual application situation.

[0182] The processor may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0183] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as program instructions / modules corresponding to the jigsaw image data processing method in the embodiment of the present application. The processor executes various functional applications and data processing of the processor by running the non-transitory software programs, instructions and modules stored in the memory, that is, realizing the jigsaw image data processing method in the above method embodiment.

[0184] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0185] One or more modules are stored in the memory, and when executed by the processor, the jigsaw image data processing method in the embodiment is executed.

[0186] In some embodiments of the present application, the user equipment may include a processor, a memory, and a transceiver unit, which may include a receiver and a transmitter. The processor, memory, receiver, and transmitter may be connected through a bus system. The memory is used to store computer instructions, and the processor is used to execute the computer instructions stored in the memory to control the transceiver unit to send and receive signals.

[0187] As an implementation method, the functions of the receiver and the transmitter in the present application can be considered to be implemented through a transceiver circuit or a dedicated chip for transceiver, and the processor can be considered to be implemented through a dedicated processing chip, a processing circuit or a general chip.

[0188] As another implementation method, it is possible to use a general-purpose computer to implement the server provided in the embodiment of the present application, that is, to store the program code for implementing the functions of the processor, receiver, and transmitter in a memory, and the general-purpose processor implements the functions of the processor, receiver, and transmitter by executing the code in the memory.

[0189] The present application also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the aforementioned puzzle image data processing method are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0190] The embodiment of the present application also provides a computer program product, which implements the aforementioned puzzle image data processing method when executed by a processor.

[0191] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is specifically performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0192] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0193] In the present application, features described and / or illustrated for one embodiment may be used in the same manner or in a similar manner in one or more other embodiments, and / or combined with features of other embodiments or replace features of other embodiments.

[0194] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for processing jigsaw puzzle image data, characterized in that: include: Get the target image time type currently corresponding to the target user account of the jigsaw puzzle game application; Obtaining a puzzle image whose image duration type is the target image duration type from each of the puzzle images as the current target puzzle image, wherein the image duration type of the puzzle image is determined after classifying the puzzle image based on content segmentation; The target jigsaw puzzle image is added to an image recommendation list corresponding to the jigsaw puzzle game application and for the target user account.

2. The jigsaw puzzle image data processing method according to claim 1, characterized in that: The method of obtaining the target image time type currently corresponding to the target user account of the jigsaw puzzle game application includes: Obtaining an average jigsaw puzzle game time corresponding to a target user account of the jigsaw puzzle game application; From the correspondence between each preset image time type and each preset time range, the image time type corresponding to the preset time range where the average time of the jigsaw puzzle game is located is determined, and the image time type has been determined as the target image time type currently corresponding to the target user account.

3. The jigsaw puzzle image data processing method according to claim 1, characterized in that: Before obtaining the target image time type currently corresponding to the target user account of the jigsaw puzzle game application, the method further includes: Preprocessing the puzzle image to delete each puzzle sub-block in the boundary area of ​​the puzzle image to obtain a corresponding initialization image; Performing image segmentation on the initialization image, and performing partition marking on each segmented area of ​​the initialization image obtained by the image segmentation; If there are different segmented areas belonging to the same content object among the segmented areas, the different segmented areas belonging to the same content object are merged to form a merged area, and a unique identifier is re-set for the merged area; The regions with the same current identifier are taken as a set, and the average feature vectors corresponding to each of the sets are extracted. If the similarity between the average feature vectors corresponding to different sets meets the similarity threshold, the sets with different similarities meeting the similarity threshold are merged into one set, and the unique identifier is reset for the merged set; Extracting features of the puzzle image according to each of the current sets to obtain feature data of the puzzle image; The image time type to which the puzzle image belongs is determined according to the feature data of the puzzle image.

4. The jigsaw puzzle image data processing method according to claim 3, characterized in that: The performing image segmentation on the initialization image and performing partition marking on each segmented area of ​​the initialization image obtained by the image segmentation includes: Input the initialization image into a preset image segmentation model so that the image segmentation model outputs corresponding segmentation areas corresponding to the initialization image; wherein the image segmentation model includes: a SAM model; Each of the partitioned areas is marked as a partition, so that each of the partitioned areas corresponds to a different unique identifier.

5. The jigsaw puzzle image data processing method according to claim 3, characterized in that: The process of taking the regions with the same current identifier as a set and extracting the average feature vectors corresponding to each of the sets, and if the similarity between the average feature vectors corresponding to different sets meets the similarity threshold, merging the sets with different similarities meeting the similarity threshold into one set, and resetting the unique identifier for the merged set, includes: The regions with the same current identifier are taken as a set, and each of the image sub-blocks in each of the sets is sequentially input into a preset image feature extraction model, so that the image feature extraction model outputs a multi-dimensional feature vector corresponding to each of the image sub-blocks; wherein the image feature extraction model includes: a Resnet model and / or a VGG model; According to the multidimensional feature vectors corresponding to the image sub-blocks in the sets, respectively, determining the average feature vectors corresponding to the sets; Determine whether the similarity between the average feature vectors corresponding to different sets satisfies the similarity threshold; if so, merge the sets whose similarities satisfy the similarity threshold into one set, and reset the unique identifier for the merged set so that each image sub-block in the same set is marked with the same unique identifier.

6. The jigsaw puzzle image data processing method according to claim 3, characterized in that: The extracting features of the puzzle image according to each of the current sets to obtain feature data of the puzzle image includes: Obtaining the number of the current sets of the puzzle image and the number of image sub-blocks corresponding to each set, wherein the number of image sub-blocks corresponding to the set is the area of ​​the set; Sorting the sets in descending order of area, and determining the mean square error of the areas of the first three sets after sorting; and determining the mean square error of the areas of all the sets of the puzzle image at present; According to the number of the sets and the mean square error of the areas of the first three sets, the proportion of the total area of ​​the first three sets is determined; and the number of the sets whose areas are less than the area threshold is determined, so as to obtain characteristic data including the number of the sets, the areas corresponding to each of the sets, the mean square error of the areas of the first three sets, the mean square error of the areas of all the sets, the proportion of the total area of ​​the first three sets and the number of the sets whose areas are less than the first area threshold.

7. The jigsaw puzzle image data processing method according to claim 6, characterized in that: The step of determining the image time type to which the puzzle image belongs according to the feature data of the puzzle image comprises: Determine whether the area of ​​the largest set among the sets corresponding to the puzzle image is equal to or greater than a second area threshold, and if so, determine that the image time type to which the puzzle image belongs is: the first type; wherein the second area threshold is greater than the first area threshold; If the area of ​​the largest set among the sets corresponding to the puzzle image is smaller than the second area threshold, determining whether the total area ratio of the first three sets corresponding to the puzzle image is equal to or greater than the ratio threshold; If the total area ratio of the first three sets corresponding to the puzzle image is equal to or greater than the ratio threshold, then determine whether the feature data corresponding to the puzzle image meets a preset condition, wherein the preset condition includes: the number of sets corresponding to the puzzle image is less than a first number threshold, the area of ​​the set with the largest area is less than a third area threshold, and the mean square deviation of the areas of all the sets is less than a mean square deviation threshold; If the feature data corresponding to the puzzle image does not meet the preset condition, the image time type to which the puzzle image belongs is determined to be the second type; if the feature data corresponding to the puzzle image meets the preset condition, the image time type to which the puzzle image belongs is determined to be the third type; If the total area ratio of the first three sets corresponding to the puzzle image is less than the ratio threshold, then determining whether the number of sets corresponding to the puzzle image whose area is less than the first area threshold exceeds a second number threshold, wherein the second number threshold is greater than the first number threshold; if so, determining that the image time type to which the puzzle image belongs is: the fourth type; Among them, the preset time ranges corresponding to the first type, the second type, the third type to the fourth type respectively decrease in sequence.

8. A jigsaw puzzle image data processing device, characterized in that: include: An image time type acquisition module is used to acquire the target image time type currently corresponding to the target user account of the jigsaw game application; a puzzle image acquisition module, used for acquiring a puzzle image whose image duration type is the target image duration type from each puzzle image, as the current target puzzle image, wherein the image duration type of the puzzle image is determined after classifying the puzzle image based on content segmentation; The puzzle image recommendation module is used to add the target puzzle image to the image recommendation list for the target user account corresponding to the puzzle game application.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for processing jigsaw image data according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for processing jigsaw image data as claimed in any one of claims 1 to 7 is implemented.

Images