Image cheating detection method, device, electronic device, and storage medium

Abstract

The present invention provides a method, device, electronic device, and storage medium for detecting cheating in an image. The method includes: obtaining multiple images of an object to be detected; classifying the multiple images to divide them into multiple image groups; for each image group, selecting two images in turn as the first image and the second image for cheating detection: detecting key points of the first image and the second image to obtain a first key point set and a second key point set, and extracting feature vectors; establishing an association relationship between the key point sets of the first key point set and the second key point set; calculating the matching rate between the first image and the second image; determining whether the matching rate between the first image and the second image is greater than a predetermined threshold; if not, determining that cheating behavior exists between the first image and the second image; if so, performing a horizontal mirror transformation on the second image, updating the transformed image to the second image, and performing cheating detection again. The method and device provided by the present invention realize image cheating detection.

Description

Background Art

[0002] At present, pictures are still the most important means for the OTA (Online Travel Agency) industry to display products. High-quality pictures help to better highlight the selling points and provide users with a better browsing experience.

[0003] OTA's various businesses need to display a large number of pictures. For example, hotel pictures need to display pictures of rooms, restaurants, swimming pools, meeting rooms and other categories. However, in actual business, there are cheating behaviors such as enlarging small pictures to pass off as good ones, and changing a single picture and displaying it repeatedly. This will lead to poor quality and redundant content of the pictures, which will affect the hotel recommendation rating and display ranking, resulting in a poor user experience. Therefore, image cheating detection has become the main means to solve this problem.

[0004] However, how to detect cheating in images in the OTA field is a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0005] In order to overcome the defects of the above-mentioned prior art, the present invention provides a method, device, electronic device and storage medium for detecting cheating in pictures, so as to realize cheating detection in pictures.

[0006] According to one aspect of the present invention, a method for detecting image cheating is provided, comprising:

[0007] S110: Acquire multiple images of the object to be detected;

[0008] S120: Classifying the multiple pictures to divide them into multiple picture groups;

[0009] S130: For each picture group, two pictures are selected in sequence as the first picture and the second picture for cheating detection:

[0010] S131: Detect key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extract feature vectors for the detected key points;

[0011] S132: establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set;

[0012] S133: calculating a matching rate between the first picture and the second picture according to an association relationship between the first key point set and the key point set of the second key point set;

[0013] S134: Determine whether a matching rate between the first picture and the second picture is greater than a predetermined threshold;

[0014] If the judgment in step S134 is no, then step S135 is executed: determining whether cheating behavior exists between the first picture and the second picture;

[0015] If the judgment in step S134 is yes, then step S136 is executed: the second image is transformed, the transformed image is updated as the second image, and steps S131 to S134 are executed again.

[0016] In some embodiments of the present invention, the step S136 includes: transforming the second image, updating the transformed image to the second image, and executing the steps S131 to S134 again.

[0017] In some embodiments of the present invention, the step S136 of transforming the second image includes: performing a horizontal mirror flip on the second image.

[0018] In some embodiments of the present invention, the step S131: detecting key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extracting feature vectors from the detected key points includes:

[0019] Construct a Gaussian pyramid for the first image and the second image respectively;

[0020] Based on the FAST algorithm, M key points are detected from the Gaussian pyramid of the first image, and the direction angle of each key point based on the grayscale centroid of the first image;

[0021] Based on the FAST algorithm, N key points are detected from the Gaussian pyramid of the second image, and the direction angle of each key point based on the grayscale centroid of the second image;

[0022] Based on the steered BRIEF feature description operator and the direction angle, feature vectors are extracted for each key point of the first image and the second image respectively.

[0023] Wherein, M and N are integers greater than 0.

[0024] In some embodiments of the present invention, the step S132 of establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set includes:

[0025] The FLANN algorithm is used to match the feature vectors of each key point of the first key point set and the second key point set, and an association relationship between the key point sets of the first key point set and the second key point set is established.

[0026] In some embodiments of the present invention, the step S133 of calculating the matching rate between the first image and the second image according to the association relationship between the first key point set and the key point set of the second key point set includes:

[0027] Acquire the number of key point pairs having an association relationship between the key point sets of the first key point set and the second key point set as the number of key point pairs;

[0028] The smaller number of key points between the first key point set and the second key point set is used as the number of key points to be calculated;

[0029] The ratio of the number of key point pairs to the number of key points to be calculated is used as the matching rate between the first image and the second image.

[0030] In some embodiments of the present invention, if it is determined that cheating occurs in the first and second images, the images with lower image performance values ​​among the first and second images are deleted from the multiple images of the object to be detected, and the sorting order of the object to be detected is adjusted based on the deleted multiple images.

[0031] According to another aspect of the present invention, there is also provided a device for detecting cheating in an image, comprising:

[0032] An acquisition module, used to acquire multiple images of the object to be detected;

[0033] A classification module, used for classifying the plurality of pictures so as to divide the pictures into a plurality of picture groups;

[0034] The detection module is used to select two pictures in each picture group as the first picture and the second picture to perform cheating detection, wherein the cheating detection includes:

[0035] S131: Detect key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extract feature vectors for the detected key points;

[0036] S132: establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set;

[0037] S133: calculating a matching rate between the first picture and the second picture according to an association relationship between the first key point set and the key point set of the second key point set;

[0038] S134: Determine whether a matching rate between the first picture and the second picture is greater than a predetermined threshold;

[0039] If the judgment in step S134 is no, then step S135 is executed: determining whether cheating behavior exists between the first picture and the second picture;

[0040] If the judgment in step S134 is yes, then step S136 is executed: the second image is transformed, the transformed image is updated as the second image, and steps S131 to S134 are executed again.

[0041] According to another aspect of the present invention, there is also provided an electronic device, comprising: a processor; and a storage medium storing a computer program, wherein the computer program executes the steps of the above-mentioned image cheating detection method when executed by the processor.

[0042] According to another aspect of the present invention, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned image cheating detection method are executed.

[0043] Compared with the prior art, the advantages of the present invention are:

[0044] The present invention proposes to classify pictures of objects such as hotels and scenic spots to be detected, and then compare them in pairs according to categories. Key point detection, feature vector extraction, feature matching and matching degree calculation are performed on each pair of pictures to detect the picture pairs with cheating behavior and delete the cheating pictures, thereby improving the quality of the displayed picture content. Thus, the detection of picture cheating behavior in production business is realized by a method based on calculating picture similarity; at the same time, by using feature point matching technology and optimizing the processing flow, not only can simple cheating behaviors such as scaling and stretching transformation be detected, but also cheating behaviors such as cutting and mirroring transformation have good effects. Furthermore, for the cheating behaviors of repeated uploading and passing inferior products off as good ones in the existing picture information of objects such as hotels, an algorithm flow is established to detect cheating behaviors by scaling, stretching, mirroring and cutting, and the comparison efficiency is improved by classifying the picture content and comparing in groups. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

[0046] Figure 1 A flow chart of a method for detecting image cheating according to an embodiment of the present invention is shown.

[0047] Figure 2 A flowchart of transforming the second picture and updating the transformed picture to the second picture according to a specific embodiment of the present invention is shown.

[0048] Figure 3A flowchart is shown of detecting key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extracting feature vectors from the detected key points according to an embodiment of the present invention.

[0049] Figure 4 A flow chart of calculating a matching rate between the first picture and the second picture according to an association relationship between the first key point set and the key point set of the second key point set is shown according to an embodiment of the present invention.

[0050] Figure 5 A module diagram of a device for detecting image cheating according to an embodiment of the present invention is shown.

[0051] Figure 6 A schematic diagram of a computer-readable storage medium in an exemplary embodiment of the present disclosure is schematically shown.

[0052] Figure 7 A schematic diagram of an electronic device in an exemplary embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0053] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the disclosure will be more comprehensive and complete and to fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0054] In addition, the accompanying drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0055] In order to solve the defects of the prior art and realize cheating detection of pictures, the present invention provides a method, device, electronic device and storage medium for detecting cheating of pictures.

[0056] See first Figure 1 , Figure 1 A schematic diagram of a method for detecting image cheating according to an embodiment of the present invention is shown. The method for detecting image cheating includes the following steps:

[0057] S110: Acquire multiple images of the object to be detected.

[0058] Specifically, the object to be detected in the OTA field may be, for example, a tourist attraction, a hotel, etc., but the present invention is not limited thereto.

[0059] S120: Classify the multiple pictures to divide them into multiple picture groups.

[0060] Specifically, a machine learning or deep learning method such as a trained neural network model may be used to classify multiple images. Taking a hotel as an example, the images may be divided into a lobby group, a guest room group, a conference room group, etc., but the present invention is not limited thereto.

[0061] S130: For each picture group, two pictures are selected in sequence as the first picture and the second picture for cheating detection.

[0062] S131: Detect key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extract feature vectors for the detected key points.

[0063] S132: Establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set.

[0064] S133: Calculate a matching rate between the first picture and the second picture according to the association relationship between the first key point set and the key point set of the second key point set.

[0065] S134: Determine whether a matching rate between the first picture and the second picture is greater than a predetermined threshold.

[0066] Specifically, the predetermined threshold value may be determined according to demand, and the present invention is not limited to this.

[0067] If the judgment in step S134 is no, then step S135 is executed: determining whether cheating behavior exists in the first picture and the second picture.

[0068] If the judgment in step S134 is yes, then step S136 is executed: the second picture is transformed, the transformed picture is updated as the second picture, and steps S131 to S134 are executed again. Furthermore, in the process of executing steps S131 to S134 again, the first picture does not need to be processed, and only the feature vector of the first picture needs to be reused.

[0069] In the image cheating detection method provided by the present invention, it is proposed to classify the images of the objects to be detected, such as hotels and scenic spots, and then compare them in pairs according to the categories. Key point detection, feature vector extraction, feature matching and matching degree calculation are performed on each pair of images to detect the image pairs with cheating behavior and delete the cheating images, so as to improve the quality of the displayed image content. Thus, the detection of image cheating behavior in the production business is realized by a method based on calculating the similarity of images; at the same time, by using the feature point matching technology and optimizing the processing flow, not only can simple cheating behaviors such as scaling and pulling transformation be detected, but also cheating behaviors such as cutting and mirroring transformation have good effects. Further, for the cheating behaviors of repeated uploading and passing inferior products off as good ones in the existing image information of objects such as hotels, an algorithm flow is established to detect cheating behaviors by scaling, pulling, mirroring and cutting, and the comparison efficiency is improved by classifying the image content and comparing in groups.

[0070] See below Figure 2 , Figure 2 FIG. 1 is a flowchart showing how to transform the second picture and update the transformed picture to the second picture according to a specific embodiment of the present invention. Specifically, Figure 2 In the step S136, the step S136 includes:

[0071] Step S1361: determining whether the number of transformations of the second image is greater than a predetermined number;

[0072] In this embodiment, the predetermined number of times is 1 time.

[0073] If the judgment in step S1361 is yes, then step S1362 is executed: judging that there is no cheating behavior between the first picture and the second picture;

[0074] If the judgment result of step S1361 is no, then step S1363 is executed: the second image is transformed, the transformed image is updated as the second image, and steps S131 to S134 are executed again.

[0075] The transforming of the second image may include any one of the following transforming methods: performing a horizontal mirror flipping of the second image.

[0076] In a preferred embodiment of the present invention, the stretching and rotation matching of the image can be implemented by step S131 and step S132. Therefore, the second image of step S136 can only include the horizontal mirror image change of the image. At this time, according to the horizontal mirror image change mode, the predetermined number of times of step S1361 can be set to 1. In this way, the system calculation amount is reduced and the system detection efficiency is improved.

[0077] See below Figure 3 , Figure 3 The flowchart of detecting key points of the first picture and the second picture to obtain a first key point set and a second key point set and extracting feature vectors from the detected key points according to an embodiment of the present invention is shown. Figure 3 In the step S131, the step S131 includes:

[0078] Step S1311: constructing a Gaussian pyramid with 8 layers by downsampling the first image and the second image at a ratio of 1 / 1.2, and detecting multi-scale key points on the pyramid to ensure scaling robustness of the extracted features;

[0079] Step S1312: Detect M key points from the Gaussian pyramid of the first image based on the FAST algorithm and the direction angle of each key point based on the grayscale centroid of the first image.

[0080] Step S1313: Detect N key points from the Gaussian pyramid of the second image based on the FAST algorithm and the direction angle of each key point based on the grayscale centroid of the second image.

[0081] Therefore, based on the FAST algorithm, M and N key points can be detected from the Gaussian pyramids of the first image IA and the second image IB, respectively, which are represented as the first key point set XA = {x a1 , x a2 , …, x am , …, x aM} and the second key point set XB = {x b1 , x b2 , …, x bn , …, x bN}, respectively based on I A and I B The grayscale centroid of the key point calculates the direction angle (main direction) Θ A ={θ a1 ,θ a2 ,…,θ am ,…,θ aM} and Θ B ={θ b1 ,θ b2 ,…,θ bn ,…,θ bN}.

[0082] Step S1314: extracting feature vectors for each key point of the first image and the second image based on the steered BRIEF feature description operator and the direction angle, where M and N are integers greater than 0.

[0083] Therefore, based on the steered BRIEF feature description operator, the first key point set X Aand the second keypoint set X B Each key point in extracts a 256-dimensional feature vector, represented as W A ={w a1 , w a2 ,…,w am ,,…,w aM} and W B ={w b1 , w b2 ,…,w bn ,,…,w bN}, thereby achieving feature extraction in the local neighborhood based on the key point position.

[0084] In some embodiments of the present invention, Figure 1 Step S132 shown: establishing an association relationship between the key point sets of the first key point set and the second key point set based on the feature vectors of each key point of the first key point set and the second key point set includes: using the FLANN algorithm to match the feature vectors of each key point of the first key point set and the second key point set, and establishing an association relationship between the key point sets of the first key point set and the second key point set.

[0085] Specifically, we can use the open source library, based on OpenCV's FLANN matcher, to quickly and efficiently match WA and WB to obtain two point sets X A and X B The corresponding relationship Corr, Corr = {(x am1 , x bn1 ), (x am2 , x bn1 ),…,(x amk , x bnk ),…,(x amK , x bnK )}, where x amk From point set X A , x bnk From point set X B , and K≤min{M,N}.

[0086] See below Figure 4 , Figure 4 A flow chart of calculating the matching rate between the first image and the second image according to the association relationship between the first key point set and the key point set of the second key point set according to an embodiment of the present invention is shown. Figure 4 In the step S133, the step S133 includes:

[0087] Step S1331: obtaining the number of key point pairs having an association relationship between the key point sets of the first key point set and the second key point set as the number of key point pairs;

[0088] Step S1332: taking the smaller number of key points in the first key point set and the second key point set as the number of key points to be calculated;

[0089] Step S1333: taking the ratio of the number of the key point pairs to the number of the key points to be calculated as the matching rate between the first picture and the second picture.

[0090] Specifically, we can calculate the similarity metric of two images based on the results of the FLANN matcher. Here we use the matching rate, which is the ratio of the number of successfully matched point pairs to the number of point sets X between the two points. A and X B The smaller value of the number of key points is calculated as follows:

[0091]

[0092] The value range of MatchRate is [0, 1], which is generally much smaller than 1. K is the number of successfully matched point pairs, M and N are the number of points in the point set X A and point set X B The number of key points in .

[0093] Further, in a specific implementation of the present invention, when the step S134 is judged to be yes, the second image IB can be flipped left and right by using the image processing interface of OpenCV to obtain I B′ Then, construct I B′ Based on the Gaussian pyramid of IB′, N′ key points are detected from the Gaussian pyramid of IB′ based on the FAST key point detection operator. The detected key point set is recorded as X B′ ={x b′1 , x b′2 , …, x b′n′ , …, x b′N′}, based on I B′ The grayscale centroid of the key point is used to calculate the main direction Θ B′ ={θ b′1 ,θ b′2 ,…,θ b′n′ ,…,θ b′N′}. N′ is an integer greater than 0. Using the steered BRIEF feature operator, for the point set X B′ Each key point in extracts a 256-dimensional feature vector, denoted as W B′ ={w b′1 , w b′2 ,…,w b′n′ ,…,w b′N′}. Using the FLANN matcher in the OpenCV open source image processing library, W Aand W B′ Match and get two point sets X A and X B′ The corresponding relationship Corr′, Corr′={(x am1 , x bn′1 ), (x am2 , x bn′2 ),…,(x amk′ , x bn′k′ ),…,(x amK′ , x bn′K′ )}, where x amk′ From point set X A , x bn′k′ From point set X B′ , and K′≤min{M,N′}.

[0094] Then, based on the results of the FLANN matcher, the two images I are calculated A and I B′ Similarity measure, that is, feature point set X A And the feature point set X B′ Match rate:

[0095]

[0096] The value range of MatchRate′ is [0, 1], which is generally much smaller than 1. K′ is the number of successfully matched point pairs, and M and N′ are the number of point sets X A and point set X B′ The number of key points in .

[0097] Further, in some embodiments of the present invention, if it is determined that cheating exists in the first and second images, the images with lower image performance values ​​(e.g., lower resolution, lower image size, lower definition of the target object in the image, or one or more) in the first and second images are deleted from the multiple images of the object to be detected, and the sorting order of the object to be detected is adjusted based on the deleted multiple images. Specifically, in this embodiment, each object to be detected can be sorted based on at least the weighted sum of the performance values ​​of its multiple images.

[0098] Therefore, an automated method process for detecting cheating behavior can be implemented based on open source image processing algorithm technology and a self-developed image classification model for the object to be detected. The comparison is based on the grouping of the object to which the image belongs and the category label, which greatly reduces the amount of calculation and is more conducive to parallel processing. By calculating the image similarity using the image feature point matching method and optimizing the process, not only simple scaling and lifting transformations can be detected, but also mirror transformations and situations with large-scale cropping can be identified. It has a wide range of applications and can significantly save operating and maintenance costs, effectively improving the operational efficiency of hotel image-related businesses in OTA scenarios.

[0099] The above is merely an illustrative description of multiple implementations of the present invention, and the present invention is not limited thereto.

[0100] The present invention also provides a device for detecting cheating in pictures. Figure 5 The schematic diagram of the image cheating detection device according to the embodiment of the present invention is shown. The image cheating detection device 200 includes an acquisition module 210 , a division module 220 and a detection module 230 .

[0101] The acquisition module 210 is used to acquire multiple images of the object to be detected.

[0102] The division module 220 is used to classify the multiple pictures to divide them into multiple picture groups.

[0103] The detection module 230 is used to select two pictures in each picture group as the first picture and the second picture to perform cheating detection, and the cheating detection includes:

[0104] S131: Detect key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extract feature vectors for the detected key points;

[0105] S132: establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set;

[0106] S133: calculating a matching rate between the first picture and the second picture according to an association relationship between the first key point set and the key point set of the second key point set;

[0107] S134: Determine whether a matching rate between the first picture and the second picture is greater than a predetermined threshold;

[0108] If the judgment in step S134 is no, then step S135 is executed: determining whether cheating behavior exists between the first picture and the second picture;

[0109] If the judgment in step S134 is yes, then step S136 is executed: the second image is transformed, the transformed image is updated as the second image, and steps S131 to S134 are executed again.

[0110] In the image cheating detection device provided by the present invention, it is proposed to classify the images of the objects to be detected, such as hotels and scenic spots, and then compare them in pairs according to the categories. Key point detection, feature vector extraction, feature matching and matching degree calculation are performed on each pair of images to detect the image pairs with cheating behavior and delete the cheating images, so as to improve the quality of the displayed image content. Thus, the detection of image cheating behavior in production business is realized by a method based on calculating the similarity of images; at the same time, by using the feature point matching technology and optimizing the processing flow, not only can simple cheating behaviors such as scaling and stretching transformation be detected, but also cheating behaviors such as cutting and mirroring transformation have good effects. Further, for the existing cheating behaviors such as repeated uploading and inferior cheating behaviors in the image information of objects such as hotels, an algorithm flow is established to detect cheating behaviors by scaling, stretching, mirroring and cutting, and improve the comparison efficiency by classifying the image content and comparing in groups.

[0111] Figure 5 The image cheating detection device provided by the present invention is only schematically shown. Without violating the concept of the present invention, the splitting, merging and adding of modules are all within the protection scope of the present invention. The image cheating detection device provided by the present invention can be implemented by software, hardware, firmware, plug-in and any combination thereof, and the present invention is not limited thereto.

[0112] In an exemplary embodiment of the present disclosure, a computer-readable storage medium is also provided, on which a computer program is stored, and when the program is executed by, for example, a processor, the steps of the image cheating detection method described in any of the above embodiments can be implemented. In some possible implementations, various aspects of the present invention can also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of various exemplary embodiments of the present invention described in the above-mentioned image cheating detection method section of this specification.

[0113] refer to Figure 6As shown, a program product 400 for implementing the above method according to an embodiment of the present invention is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto. In this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0114] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0115] The computer readable storage medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The readable storage medium may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by an instruction execution system, an apparatus, or a device or used in combination with it. The program code contained on the readable storage medium may be transmitted with any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0116] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the tenant computing device, partially on the tenant device, as a separate software package, partially on the tenant computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device may be connected to the tenant computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0117] In an exemplary embodiment of the present disclosure, an electronic device is also provided, which may include a processor and a memory for storing executable instructions of the processor, wherein the processor is configured to execute the steps of the image cheating detection method described in any one of the above embodiments by executing the executable instructions.

[0118] It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as a system, method or program product. Therefore, various aspects of the present invention may be specifically implemented in the following forms, namely: a complete hardware implementation, a complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which may be collectively referred to herein as a "circuit", "module" or "system".

[0119] Refer to the following Figure 7 The electronic device 600 according to this embodiment of the present invention is described. Figure 7 The electronic device 600 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0120] like Figure 7 As shown, the electronic device 600 is in the form of a general computing device. The components of the electronic device 600 may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different system components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0121] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 executes the steps of various exemplary embodiments of the present invention described in the above-mentioned image cheating detection method section of this specification. For example, the processing unit 610 can execute the following steps: Figures 1 to 3 Follow the steps shown in .

[0122] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 6201 and / or a cache memory unit 6202 , and may further include a read-only memory unit (ROM) 6203 .

[0123] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules and program data, each of which or some combination may include the implementation of a network environment.

[0124] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0125] The electronic device 600 may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may communicate with one or more devices that enable tenants to interact with the electronic device 600, and / or may communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed through an input / output (I / O) interface 650. In addition, the electronic device 600 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 through a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0126] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, or a network device, etc.) to execute the above-mentioned picture cheating detection method according to the implementation of the present disclosure.

[0127] Compared with the prior art, the advantages of the present invention are:

[0128] The present invention proposes to classify pictures of objects such as hotels and scenic spots to be detected, and then compare them in pairs according to categories. Key point detection, feature vector extraction, feature matching and matching degree calculation are performed on each pair of pictures to detect the picture pairs with cheating behavior and delete the cheating pictures, thereby improving the quality of the displayed picture content. Thus, the detection of picture cheating behavior in production business is realized by a method based on calculating picture similarity; at the same time, by using feature point matching technology and optimizing the processing flow, not only can simple cheating behaviors such as scaling and stretching transformation be detected, but also cheating behaviors such as cutting and mirroring transformation have good effects. Furthermore, for the cheating behaviors of repeated uploading and passing inferior products off as good ones in the existing picture information of objects such as hotels, an algorithm flow is established to detect cheating behaviors by scaling, stretching, mirroring and cutting, and the comparison efficiency is improved by classifying the picture content and comparing in groups.

[0129] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the appended claims.

Claims

1. A method for detecting image cheating, characterized in that: include: S110: Acquire multiple images of the object to be detected; S120: Classifying the multiple pictures to divide them into multiple picture groups; S130: For each picture group, two pictures are selected in sequence as the first picture and the second picture for cheating detection: S131: Detect key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extract feature vectors according to the detected key points; S132: establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set; S133: calculating a matching rate between the first picture and the second picture according to an association relationship between the first key point set and the key point set of the second key point set; S134: Determine whether a matching rate between the first picture and the second picture is greater than a predetermined threshold; If the judgment in step S134 is no, then step S135 is executed: determining whether cheating behavior exists between the first picture and the second picture; If the judgment in step S134 is yes, then executing step S136: transforming the second picture, updating the transformed picture as the second picture, and executing steps S131 to S134 again; The step S133: calculating the matching rate between the first picture and the second picture according to the association relationship between the key point sets of the first key point set and the second key point set includes: obtaining the number of key point pairs having an association relationship between the key point sets of the first key point set and the second key point set as the number of key point pairs; taking the smaller number of key points between the number of key points of the first key point set and the number of key points of the second key point set as the number of key points to be calculated; and taking the ratio of the number of key point pairs to the number of key points to be calculated as the matching rate between the first picture and the second picture; The matching rate is the smaller value of the number of key point pairs that are successfully matched than the number of key points in the first key point set and the second key point set, and is calculated as follows: MatchRate is the matching rate, K is the number of successfully matched point pairs, and M and N are the numbers of key points in the first key point set and the second key point set, respectively.

2. The image cheating detection method according to claim 1, characterized in that: The step S136 includes: transforming the second picture, updating the transformed picture to the second picture, and executing the steps S131 to S134 again.

3. The image cheating detection method according to claim 1, characterized in that: The step S136 of transforming the second image includes horizontally mirror-flipping the second image.

4. The image cheating detection method according to claim 1, characterized in that: The step S131: detecting key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extracting feature vectors according to the detected key points includes: Construct a Gaussian pyramid for the first image and the second image respectively; Based on the FAST algorithm, M key points are detected from the Gaussian pyramid of the first image, and the direction angle of each key point based on the grayscale centroid of the first image; Based on the FAST algorithm, N key points are detected from the Gaussian pyramid of the second image, and the direction angle of each key point based on the grayscale centroid of the second image; Based on the steered BRIEF feature description operator and the direction angle, feature vectors are extracted for each key point of the first image and the second image respectively. Wherein, M and N are integers greater than 0.

5. The image cheating detection method according to claim 1, characterized in that: The step S132: establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set includes: The FLANN algorithm is used to match the feature vectors of each key point of the first key point set and the second key point set, and an association relationship between the key point sets of the first key point set and the second key point set is established.

6. The image cheating detection method according to claim 1, characterized in that: If it is determined that cheating exists in the first picture and the second picture, the picture with a lower picture performance value in the first picture and the second picture is deleted from the multiple pictures of the object to be detected, and the sorting order of the object to be detected is adjusted based on the deleted multiple pictures.

7. A device for detecting cheating in an image, characterized in that: include: An acquisition module, used to acquire multiple images of the object to be detected; A classification module, used for classifying the plurality of pictures so as to divide the pictures into a plurality of picture groups; The detection module is used to select two pictures in each picture group as the first picture and the second picture to perform cheating detection, wherein the cheating detection includes: S131: Detect key points of the first picture and the second picture to obtain a first key point set and a second key point set, and extract feature vectors for the detected key points; S132: establishing an association relationship between the key point sets of the first key point set and the second key point set according to the feature vectors of the key points of the first key point set and the second key point set; S133: calculating a matching rate between the first picture and the second picture according to an association relationship between the first key point set and the key point set of the second key point set; S134: Determine whether a matching rate between the first picture and the second picture is greater than a predetermined threshold; If the judgment in step S134 is no, then step S135 is executed: determining whether cheating behavior exists between the first picture and the second picture; If the judgment in step S134 is yes, then executing step S136: transforming the second picture, updating the transformed picture as the second picture, and executing steps S131 to S134 again; The step S133: calculating the matching rate between the first picture and the second picture according to the association relationship between the key point sets of the first key point set and the second key point set includes: obtaining the number of key point pairs having an association relationship between the key point sets of the first key point set and the second key point set as the number of key point pairs; taking the smaller number of key points between the number of key points of the first key point set and the number of key points of the second key point set as the number of key points to be calculated; and taking the ratio of the number of key point pairs to the number of key points to be calculated as the matching rate between the first picture and the second picture; The matching rate is the smaller value of the number of key point pairs that are successfully matched than the number of key points in the first key point set and the second key point set, and is calculated as follows: MatchRate is the matching rate, K is the number of successfully matched point pairs, and M and N are the numbers of key points in the first key point set and the second key point set, respectively.

8. An electronic device, characterized in that: The electronic device comprises: processor; A storage medium having a computer program stored thereon, wherein the computer program, when executed by the processor, executes the image cheating detection method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the image cheating detection method according to any one of claims 1 to 6 is executed.

Images