Image tampering detection method, device, equipment and storage medium
By performing pre-tampering region screening and feature extraction on images, and combining deep learning networks and the SIFT algorithm, the location of image tampering can be quickly and accurately located, solving the balance problem between detection speed and accuracy in existing technologies.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- MACAO POLYTECHNIC INST
- Filing Date
- 2024-01-25
- Publication Date
- 2026-05-19
AI Technical Summary
Existing technologies struggle to simultaneously achieve rapid detection and accurate location of image tampering.
By performing image detection on the image to be detected, the pre-tampered region is identified, tampering screening and feature extraction are carried out, and multi-scale feature extraction and feature matching are performed using deep learning networks and SIFT algorithms to locate the image tampered region.
This improves the speed of image tampering detection while ensuring accurate localization of the tampered area.
Smart Images
Figure CN117935033B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of image detection technology, and in particular to an image tampering detection method, apparatus, device, and storage medium. Background Technology
[0002] Image tampering detection plays a crucial role in several key areas. In legal and court settings, it verifies the authenticity of digital evidence, ensuring it remains unaltered. In news and media reporting, the technology helps detect and prevent image tampering, ensuring viewers receive accurate information. In advertising and marketing, image tampering detection protects advertisements from malicious alteration, ensuring the credibility of advertising content. On social media platforms, it prevents the spread and misleading dissemination of misinformation. In identity verification systems, image tampering detection ensures the authenticity of images such as ID photos to prevent fraud. Furthermore, it plays a critical role in the preservation of art and cultural relics, the integrity of medical images, and academic research, maintaining the authenticity and integrity of images and ensuring credibility and security across various fields.
[0003] Currently, passive tamper detection methods are mainly divided into two categories: one is tamper detection based on deep learning networks, and the other is based on feature points. Feature-point-based detection schemes show high accuracy, but their computation time is long. Conversely, deep network schemes can quickly locate tampered regions in images and have stronger robustness, but their accuracy is significantly lower than that of feature-point-based methods. The urgent problem we need to solve is how to achieve a balance between accuracy and speed.
[0004] Therefore, there is an urgent need for an image tampering detection method that can quickly detect whether an image has been tampered with and accurately locate the location of the tampering.
[0005] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention
[0006] The main objective of this invention is to provide an image tampering detection method, apparatus, device, and storage medium, aiming to solve the technical problem of how to quickly and accurately locate the image tampering position in the prior art.
[0007] To achieve the above objectives, the present invention provides an image tampering detection method, the method comprising the following steps:
[0008] Image detection is performed on the image to be detected to obtain the results of the pre-tampered region;
[0009] Based on the pre-tampered region results, tampering filtering is performed to obtain the target tampered region results;
[0010] The tampering features of the target tampered area are extracted to obtain image tampering feature information;
[0011] Based on the image tampering feature information, the tampered area is located to obtain the image tampering detection result.
[0012] Optionally, image detection is performed on the image to be detected to obtain the pre-tampered region result. Specific steps include:
[0013] Image features are extracted from the image to be detected to obtain the feature image to be detected;
[0014] Based on the feature image to be detected, perform feature similarity calculation to obtain the relevant image to be detected;
[0015] Multi-scale feature extraction is performed on the image to be detected to obtain the multi-scale features to be detected;
[0016] Image tampering prediction is performed based on the relevant image to be detected and the multi-scale features to be detected to obtain the pre-tampered region result.
[0017] Optionally, feature similarity calculation is performed on the feature image to be detected to obtain the relevant image to be detected. Specific steps include:
[0018] Based on the feature image to be detected, a feature similarity calculation is performed to obtain the feature similarity result. The feature similarity calculation formula is as follows:
[0019]
[0020] In the formula, Represents the eigenvector f_v i and f_v j The similarity; i and j represent different spatial coordinates in the feature vector; W is the width of the image to be detected; H is the height of the image to be detected;
[0021] The images to be detected are obtained by ranking the similarity results based on feature similarity.
[0022] Optionally, based on the pre-tampered region results, tampering filtering is performed to obtain the target tampered region results. Specific steps include:
[0023] The pre-tampered region result is masked to obtain a masked filtering image;
[0024] The image to be detected is tampered with based on the mask filtering map to obtain the target tampered area result.
[0025] Optionally, the tampered region is located based on the image tampering feature information to obtain the image tampering detection result. Specific steps include:
[0026] Based on the image tampering feature information, image feature point matching is performed to obtain the feature point matching result;
[0027] Based on the feature point matching results, a displacement transformation is performed on the target tampered region to obtain a tampered displacement map;
[0028] Based on the target tampered region result and the tampered displacement map, the tampered region is located to obtain the image tampering detection result.
[0029] Optionally, image feature point matching is performed based on the image tampering feature information to obtain the feature point matching result. Specific steps include:
[0030] Based on the image tampering feature information, feature point matching and filtering are performed to obtain a set of feature points to be matched;
[0031] Each feature point in the set of feature points to be matched is integrated and matched to obtain the first feature point matching result;
[0032] The first feature point matching result is filtered out by a matching result filtering algorithm to obtain the feature point matching result.
[0033] Optionally, the tampered region is located based on the target tampered region result and the tampered displacement map to obtain the image tampering detection result. Specific steps include:
[0034] The image tampering detection result is obtained by calculating the similarity between each pixel point between the target tampered region result and the tampered displacement map using a tampering localization algorithm.
[0035] Furthermore, to achieve the above objectives, the present invention also proposes an image tampering detection device, the image tampering detection device comprising:
[0036] Image detection module: Performs image detection on the image to be detected to obtain the result of the pre-tampered region;
[0037] Detection and filtering module: Performs tampering filtering based on the pre-tampered area results to obtain the target tampered area results;
[0038] Feature extraction module: Extracts tampering features from the target tampered region to obtain image tampering feature information;
[0039] Tampering location module: Based on the image tampering feature information, locate the image tampering area and obtain the image tampering detection result.
[0040] Furthermore, to achieve the above objectives, the present invention also proposes an image tampering detection device, which includes: a memory, a processor, and an image tampering detection program stored in the memory and executable on the processor, wherein the image tampering detection program is configured to implement the steps of the image tampering detection method described above.
[0041] Furthermore, to achieve the above objectives, the present invention also proposes a computer-readable storage medium storing a computer program, wherein the storage medium stores an image tampering detection program, and when the image tampering detection program is executed by a processor, it implements the steps of the image tampering detection method as described above.
[0042] This invention performs image detection on the image to be detected to obtain a pre-tampered region result; based on the pre-tampered region result, it performs tampering screening to obtain a target tampered region result; it extracts tampering features from the target tampered region result to obtain image tampering feature information; and it locates the tampered region based on the image tampering feature information to obtain an image tampering detection result. This invention first determines the pre-tampered region of the image to be detected, and then performs subsequent image tampering detection on it to obtain the corresponding tampering detection result, thereby improving the image tampering detection speed while ensuring the accuracy of image tampering region location. Attached Figure Description
[0043] Figure 1 This is a schematic diagram of the structure of the image tampering detection device in the hardware operating environment involved in the embodiments of the present invention;
[0044] Figure 2 This is a flowchart illustrating the first embodiment of the image tampering detection method of the present invention;
[0045] Figure 3 This is a flowchart illustrating the second embodiment of the image tampering detection method of the present invention;
[0046] Figure 4 This is a flowchart illustrating the third embodiment of the image tampering detection method of the present invention;
[0047] Figure 5 This is a flowchart illustrating the fourth embodiment of the image tampering detection method of the present invention;
[0048] Figure 6 This is a structural block diagram of the first embodiment of the image tampering detection device of the present invention.
[0049] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0050] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.
[0051] Reference Figure 1 , Figure 1 This is a schematic diagram of the image tampering detection device structure in the hardware operating environment involved in the embodiments of the present invention.
[0052] like Figure 1 As shown, the image tampering detection device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wireless-Fidelity (Wi-Fi) interface). The memory 1005 may be high-speed random access memory (RAM) or stable non-volatile memory (NVM), such as a disk storage device. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.
[0053] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on the image tampering detection device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.
[0054] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a network communication module, a user interface module, and an image tampering detection program.
[0055] exist Figure 1 In the image tampering detection device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the image tampering detection device of the present invention can be set in the image tampering detection device, and the image tampering detection device calls the image tampering detection program stored in the memory 1005 through the processor 1001 and executes the image tampering detection method provided in the embodiment of the present invention.
[0056] This invention provides an image tampering detection method, referring to... Figure 2 , Figure 2This is a flowchart illustrating the first embodiment of an image tampering detection method according to the present invention.
[0057] In this embodiment, the image tampering detection method includes the following steps:
[0058] Step S10: Perform image detection on the image to be detected to obtain the pre-tampered region result;
[0059] It should be noted that, in the specific implementation, image detection of the image to be detected is achieved through a correlation-based deep learning network. This network consists of four functional modules: a feature extraction module, a correlation module, a multi-scale feature fusion module, and a prediction module. The specific image detection process is as follows: first, the feature extraction module extracts the feature image of the image to be detected; then, the correlation module and the multi-scale feature fusion module respectively obtain a similarity map and extract multi-scale features from the feature image; finally, based on the obtained similarity map and multi-scale features, the prediction module determines the pre-tampering region of the image to be detected, obtaining the corresponding pre-tampering region result.
[0060] It should also be noted that the pre-tampered area result is a region of the image to be detected that may have been tampered with, that is, the region of the image to be detected that may have been tampered with.
[0061] Step S20: Perform tampering filtering based on the pre-tampering region results to obtain the target tampering region results;
[0062] It is understandable that the target tampering area result specifically refers to the area on the image to be detected that has been tampered with.
[0063] Step S30: Extract tampering features from the target tampered area to obtain image tampering feature information;
[0064] It should be noted that, in the specific implementation, the extraction of tampering features from the target tampered region is achieved through the SIFT (Scale invariant feature transform) algorithm. Since the descriptors generated by the SIFT algorithm exhibit strong stability in terms of scale and rotation invariance, the obtained image tampering feature information can improve the accuracy of tampering location when used for tampering region localization.
[0065] Step S40: Locate the tampered area based on the image tampering feature information to obtain the image tampering detection result.
[0066] It should be noted that the image tampering detection results specifically include the range information of the tampered area and the corresponding specific location information.
[0067] This embodiment performs image detection on the image to be detected to obtain a pre-tampered region result; based on the pre-tampered region result, tampering filtering is performed to obtain a target tampered region result; tampering features are extracted from the target tampered region result to obtain image tampering feature information; and tampered region location is performed based on the image tampering feature information to obtain an image tampering detection result. This embodiment first determines the pre-tampered region of the image to be detected, and then performs subsequent image tampering detection on it to obtain the corresponding tampering detection result, thereby improving the image tampering detection speed while ensuring the accuracy of image tampering region location.
[0068] refer to Figure 3 , Figure 3 This is a flowchart illustrating a second embodiment of an image tampering detection method according to the present invention.
[0069] Based on the first embodiment described above, in this embodiment, step S10 specifically includes the following steps:
[0070] Step S11: Extract image features from the image to be detected to obtain the feature image to be detected;
[0071] Understandably, in the specific implementation, image feature extraction of the image to be detected is achieved through a feature extraction module of a correlation-based deep learning network. This feature extraction module uses VGG16 (VisualGeometry Group 16) as the backbone network, removing the last fully connected layer. Specifically, the image to be detected is processed through a series of convolutional and max-pooling layers to obtain the corresponding feature map, i.e., the feature image to be detected.
[0072] It should also be noted that in the actual implementation, as the convolution operation proceeds, the number of channels in the feature map gradually increases. At the same time, through each max pooling operation, the resolution of the feature map is reduced by half.
[0073] Step S12: Perform feature similarity calculation on the feature image to be detected to obtain the relevant image to be detected;
[0074] It should be noted that, in the specific implementation, image feature extraction of the image to be detected is achieved through the correlation module of a correlation-based deep learning network. Furthermore, since the specific objective is to detect similar regions within the image to be detected, the correlation module is needed to calculate the feature similarity of the feature images to be detected and construct the corresponding weight map.
[0075] Step S13: Perform multi-scale feature extraction on the image to be detected to obtain the multi-scale features to be detected;
[0076] It should be noted that, in the specific implementation, image feature extraction of the image to be detected is achieved through a multi-scale feature fusion module based on a correlation-based deep learning network. This multi-scale feature fusion module specifically includes ASPP (Atrous Spatial Pyramid Pooling), which is used to integrate features at different scales.
[0077] Step S14: Based on the relevant image to be detected and the multi-scale features to be detected, perform image tampering prediction to obtain the pre-tampering region result.
[0078] Understandably, in practical implementations, to ensure that the predicted result matches the size of the original image, an upsampling operation is required after each convolution operation of the dimensionality-reduced tensor. For example, by using an upsampling dilation factor of size 2 and performing 5 upsampling operations, an approximate region of the predicted tampered area in the image to be detected can be obtained, i.e., the pre-tampered region result.
[0079] This embodiment first obtains the feature image to be detected based on the image to be detected, then performs feature similarity calculation and multi-scale feature extraction on the feature image to be detected, thereby obtaining the relevant image to be detected and the multi-scale features to be detected, and finally further obtains the pre-tampered region result, realizing the approximate bounding of the tampered detection region of the image to be detected, thereby improving the tampering detection rate.
[0080] refer to Figure 4 , Figure 4 This is a flowchart illustrating a third embodiment of an image tampering detection method according to the present invention.
[0081] Based on the second embodiment described above, in this embodiment, step S12 specifically includes the following steps:
[0082] Step S121: Perform feature similarity calculation on the feature image to be detected to obtain the feature similarity result. The feature similarity calculation formula is as follows:
[0083]
[0084] In the formula, Represents the eigenvector f_v i and f_v j The similarity; i and j represent different spatial coordinates in the feature vector; W is the width of the image to be detected; H is the height of the image to be detected;
[0085] It should be noted that, in the specific implementation, the feature similarity calculation based on the feature image to be detected can ultimately yield a similarity score map between feature vectors.
[0086] ;
[0087] in, Represents any two eigenvectors and The similarity is given by i and j, where i and j represent different spatial locations in the feature map.
[0088] Step S122: Sort the images by similarity based on the feature similarity results to obtain the relevant images to be detected.
[0089] Understandably, in the actual implementation, after obtaining the similarity score map, the similarity scores are sorted from largest to smallest to select highly similar regions, i.e., the relevant images to be detected.
[0090] This embodiment calculates the feature similarity of the feature images to be detected, obtains the feature similarity results, and then sorts the similarity to obtain the relevant images to be detected, providing a data foundation for subsequent tampering location.
[0091] Further, based on the pre-tampered region result, tampering filtering is performed to obtain the target tampered region result. The specific steps include: performing mask conversion on the pre-tampered region result to obtain a mask filtering map; and performing tampering filtering on the image to be detected based on the mask filtering map to obtain the target tampered region result.
[0092] It should be noted that, in practical implementations, image masks are generally used to occlude (all or part) the image being processed to control the area or process of subsequent image processing. In this embodiment, the unoccluded area in the obtained mask filtering image is the image tampering area.
[0093] refer to Figure 5 , Figure 5 This is a flowchart illustrating the fourth embodiment of an image tampering detection method according to the present invention.
[0094] Based on the first embodiment described above, in this embodiment, step S40 specifically includes the following steps:
[0095] Step S41: Perform image feature point matching based on image tampering feature information to obtain feature point matching results;
[0096] It should be noted that in the specific implementation, each image feature point will generate a corresponding local feature descriptor through the SIFT algorithm. The specific process of matching image feature points of different image feature points is as follows: first, the nearest neighbor filtering method is used to exclude image feature points that do not have a matching relationship; then, the random sampling consensus algorithm is used to eliminate possible false matching; and finally, the zero-mean normalized cross-correlation algorithm is used to calculate the corresponding similarity of pixels between matching points at different locations.
[0097] It should also be noted that, in the specific implementation, since there is a sufficient difference between the first best matching point and the second best matching point of each image feature point, the reliability of the matching can be evaluated by calculating the difference ratio between the two matches.
[0098] Step S42: Perform displacement transformation on the target tampered area result based on the feature point matching result to obtain the tampered displacement map;
[0099] Understandably, in the specific implementation, since the matching points corresponding to any two image feature points can be obtained through geometric transformation, a corresponding homography matrix can be calculated based on the obtained matching points. Then, the mask image can be shifted using this homography matrix to obtain the tampered displacement image.
[0100] Step S43: Locate the tampered area based on the target tampered area result and the tampered displacement map to obtain the image tampering detection result.
[0101] It should be noted that, in the specific implementation, the process of locating the tampered area is to determine the similarity between the tampered displacement map and the mask map in the result of the target tampered area through the zero-mean normalized cross-correlation algorithm, thereby achieving the location of the tampered area.
[0102] In this embodiment, the corresponding feature point matching results are first obtained based on the image tampering feature information. Then, the target tampering area results are subjected to displacement transformation to obtain the tampering displacement map. Finally, the tampering area is located to obtain the image tampering detection result, thereby realizing the image tampering area location.
[0103] Further, image feature point matching is performed based on the image tampering feature information to obtain feature point matching results. The specific steps include: performing feature point matching and filtering based on the image tampering feature information to obtain a set of feature points to be matched; integrating and matching each feature point to be matched in the set of feature points to be matched to obtain a first feature point matching result; and filtering out the first feature point matching result using a matching result filtering algorithm to obtain the final feature point matching result.
[0104] It should be noted that, in the specific implementation, the matching result filtering algorithm specifically refers to the nearest neighbor filtering algorithm, which is a non-probabilistic collaborative filtering algorithm. Its specific filtering process can be as follows: First, a minimum sample set is randomly selected and used to estimate the geometric transformation model. Then, the estimated geometric transformation model is used to predict the values of all other data points, and the error between the predicted and actual values is calculated. Next, points whose errors with the model's predictions are less than a certain threshold are identified. All points that do not meet this condition are considered outliers, i.e., mismatched feature points, and these points are removed from subsequent calculations.
[0105] Further, based on the target tampered region result and the tampered displacement map, the tampered region is located to obtain the image tampering detection result. The specific steps include: calculating the similarity of each pixel between the target tampered region result and the tampered displacement map using a tampering localization algorithm to obtain the image tampering detection result.
[0106] It should be noted that, in the specific implementation, the tampering localization algorithm specifically refers to the zero-mean normalized cross-correlation algorithm. In this embodiment, the zero-mean normalized cross-correlation algorithm is used to compare the similarity of each pixel between the target tampered region result and the tampered displacement map. The specific tampering localization process can be as follows: first, calculate the mean of each image; then, perform zero-mean normalization on the images; next, calculate the similarity metric through normalized cross-correlation; finally, find the position with the highest similarity metric, which represents the region where the two images are most similar at that position.
[0107] Furthermore, embodiments of the present invention also propose a computer-readable storage medium storing a computer program, wherein the storage medium stores an image tampering detection program, and when the image tampering detection program is executed by a processor, it implements the steps of the image tampering detection method as described above.
[0108] Since this storage medium adopts all the technical solutions of all the above embodiments, it has at least all the beneficial effects brought about by the technical solutions of the above embodiments, which will not be repeated here.
[0109] Reference Figure 6 , Figure 6 This is a structural block diagram of the first embodiment of the image tampering detection device of the present invention.
[0110] like Figure 6 As shown, the image tampering detection device proposed in this embodiment of the invention includes:
[0111] Image detection module 10: Performs image detection on the image to be detected to obtain the result of the pre-tampered region;
[0112] Detection and filtering module 20: Performs tampering filtering based on the pre-tampered area results to obtain the target tampered area results;
[0113] Feature extraction module 30: Extracts tampering features from the target tampered area to obtain image tampering feature information;
[0114] Tampering location module 40: Locates the tampered area of the image based on the image tampering feature information and obtains the image tampering detection result.
[0115] This embodiment performs image detection on the image to be detected to obtain a pre-tampered region result; based on the pre-tampered region result, tampering filtering is performed to obtain a target tampered region result; tampering features are extracted from the target tampered region result to obtain image tampering feature information; and tampered region location is performed based on the image tampering feature information to obtain an image tampering detection result. This embodiment first determines the pre-tampered region of the image to be detected, and then performs subsequent image tampering detection on it to obtain the corresponding tampering detection result, thereby improving the image tampering detection speed while ensuring the accuracy of image tampering region location.
[0116] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solutions of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.
[0117] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.
[0118] In addition, for technical details not described in detail in this embodiment, please refer to the image tampering detection method provided in any embodiment of the present invention, which will not be repeated here.
[0119] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.
[0120] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.
[0121] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.
[0122] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.
Claims
1. A method for detecting image tampering, characterized in that, include: Image detection is performed on the image to be detected to obtain the results of the pre-tampered region; Based on the pre-tampered region results, tampering filtering is performed to obtain the target tampered region results; The SIFT algorithm is used to extract tampering features from the target tampered region to obtain image tampering feature information. Based on the image tampering feature information, the tampered region is located to obtain the image tampering detection result. The specific steps include: Based on the image tampering feature information, image feature point matching is performed to obtain the feature point matching result; Based on the feature point matching results, a displacement transformation is performed on the target tampered region to obtain a tampered displacement map; Based on the target tampered region result and the tampered displacement map, the tampered region is located to obtain the image tampering detection result.
2. The image tampering detection method according to claim 1, characterized in that, The image to be detected is subjected to image detection to obtain the pre-tampered region result. The specific steps include: Image features are extracted from the image to be detected to obtain the feature image to be detected; Based on the feature image to be detected, perform feature similarity calculation to obtain the relevant image to be detected; Multi-scale feature extraction is performed on the image to be detected to obtain the multi-scale features to be detected; Image tampering prediction is performed based on the relevant image to be detected and the multi-scale features to be detected to obtain the pre-tampered region result.
3. The image tampering detection method according to claim 2, characterized in that, The detection of related images involves performing feature similarity calculations on the image to be detected, and the specific steps include: Based on the feature image to be detected, a feature similarity calculation is performed to obtain the feature similarity result. The feature similarity calculation formula is as follows: ； In the formula, Represents the eigenvector f_v i and f_v j The similarity; i and j represent different spatial coordinates in the feature vector; W is the width of the image to be detected; H is the height of the image to be detected; The images to be detected are obtained by ranking the similarity results based on feature similarity.
4. The image tampering detection method according to any one of claims 1-3, characterized in that, Based on the pre-tampered region results, tampering filtering is performed to obtain the target tampered region results. The specific steps include: The pre-tampered region result is masked to obtain a masked filtering image; The image to be detected is tampered with based on the mask filtering map to obtain the target tampered area result.
5. The image tampering detection method according to claim 1, characterized in that, Based on the image tampering feature information, image feature point matching is performed to obtain the feature point matching result. The specific steps include: Based on the image tampering feature information, feature point matching and filtering are performed to obtain a set of feature points to be matched; Each feature point in the set of feature points to be matched is integrated and matched to obtain the first feature point matching result; The first feature point matching result is filtered out by a matching result filtering algorithm to obtain the feature point matching result.
6. The image tampering detection method according to claim 5, characterized in that, Based on the target tampered region result and the tampered displacement map, the tampered region is located to obtain the image tampering detection result. The specific steps include: The image tampering detection result is obtained by calculating the similarity between each pixel point between the target tampered region result and the tampered displacement map using a tampering localization algorithm.
7. An image tampering detection device, characterized in that, The image tampering detection device includes: Image detection module: Performs image detection on the image to be detected to obtain the result of the pre-tampered region; Detection and filtering module: Performs tampering filtering based on the pre-tampered area results to obtain the target tampered area results; Feature extraction module: Extracts tampering features from the target tampered region using the SIFT algorithm to obtain image tampering feature information; Tampering location module: Based on the image tampering feature information, locates the tampered area in the image and obtains the image tampering detection result, specifically including: Based on the image tampering feature information, image feature point matching is performed to obtain the feature point matching result; Based on the feature point matching results, a displacement transformation is performed on the target tampered region to obtain a tampered displacement map; Based on the target tampered region result and the tampered displacement map, the tampered region is located to obtain the image tampering detection result.
8. An image tampering detection device, characterized in that, The image tampering detection device includes: a memory, a processor, and an image tampering detection program stored in the memory and executable on the processor, wherein the image tampering detection program is configured to implement the image tampering detection method according to any one of claims 1 to 6.
9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it is able to implement the steps of the image tampering detection method according to any one of claims 1 to 6.