Image tampering detection method, device, equipment and storage medium

Abstract

The present application discloses an image tampering detection method, device, equipment and storage medium. The method comprises: obtaining an image to be detected; performing noise feature recognition on the image to be detected to obtain spatial feature information of light response non-uniform noise and frequency feature information of light response non-uniform noise in the image to be detected; inputting the spatial feature information and the frequency feature information into an image tampering detection network to perform image tampering detection, and obtaining image detection information corresponding to the image to be detected. In the scenario of image tampering detection of media content, the present application combines the spatial features and frequency features of light response non-uniform noise in an image to improve the accuracy of tampering detection results.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an image tampering detection method, device, equipment and storage medium. Background Art

[0002] In recent years, with the development of digital technology and network communication technology, the ways of modifying digital images have become diversified, such as using PhotoShop (image processing software), CrazyTalk (facial animation software) and various mobile phone editing apps. As image tampering techniques become more and more advanced and the tampering becomes more and more concealed, the difficulty of detecting image tampering continues to increase.

[0003] Although there are many image tampering detection methods, most of these methods only consider image tampering detection through a single feature, such as: tampering detection through the spatial domain characteristics of the PRNU (Photo Response Non-Uniformity) noise of the image. When the image scene content is complex and changeable, the tampering detection of a single feature is difficult to guarantee the accuracy of the detection result. Therefore, it is necessary to provide a more accurate technical solution. Summary of the invention

[0004] The present application provides an image tampering detection method, apparatus, device and storage medium, which combines the spatial domain characteristics and frequency domain characteristics of the light response non-uniform noise in the image to improve the accuracy of the tampering detection result. The technical solution of the present application is as follows:

[0005] In one aspect, a method for detecting image tampering is provided, the method comprising:

[0006] Acquire the image to be detected;

[0007] Performing noise feature recognition on the image to be detected to obtain spatial domain feature information of light response non-uniform noise in the image to be detected and frequency domain feature information of the light response non-uniform noise;

[0008] The spatial domain feature information and the frequency domain feature information are input into an image tampering detection network to perform image tampering detection, and image detection information corresponding to the image to be detected is obtained.

[0009] On the other hand, an image tampering detection device is provided, and the method includes:

[0010] An image acquisition module to be detected is used to acquire the image to be detected;

[0011] A noise feature recognition module, used to perform noise feature recognition on the image to be detected, and obtain spatial domain feature information of the light response non-uniform noise in the image to be detected and frequency domain feature information of the light response non-uniform noise;

[0012] The image tampering detection module is used to input the spatial domain feature information and the frequency domain feature information into the image tampering detection network to perform image tampering detection and obtain image detection information corresponding to the image to be detected.

[0013] On the other hand, an image tampering detection device is provided, the device comprising a processor and a memory, the memory storing at least one instruction or at least one program, the at least one instruction or the at least one program being loaded and executed by the processor to implement the image tampering detection method as described in the first aspect.

[0014] On the other hand, a computer-readable storage medium is provided, in which at least one instruction or at least one program is stored, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the image tampering detection method as described in the first aspect.

[0015] On the other hand, a computer program product or a computer program is provided, the computer program product or the computer program comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the image tampering detection method as described in the first aspect.

[0016] The present application provides an image tampering detection method, device, equipment and storage medium, which have the following technical effects:

[0017] In the scenario of image tampering detection, the present application obtains an image to be detected, and then performs noise feature recognition on the image to be detected to obtain spatial feature information of light response non-uniform noise in the image to be detected and frequency domain feature information of the light response non-uniform noise; then, the spatial feature information and the frequency domain feature information are input into an image tampering detection network to perform image tampering detection to obtain image detection information corresponding to the image to be detected. By increasing the feature dimension and utilizing the dual features of spatial and frequency domains of light response non-uniform noise to perform image tampering detection, the accuracy of image tampering detection can be greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 is a schematic diagram of an application environment provided by an embodiment of the present application;

[0020] Figure 2 It is a flowchart of an image tampering detection method provided in an embodiment of the present application;

[0021] Figure 3 It is a schematic diagram of a process of obtaining an image to be detected provided by an embodiment of the present application;

[0022] Figure 4 It is a flow chart of performing noise feature recognition on an image to be detected and obtaining spatial domain feature information and frequency domain feature information of light response non-uniform noise in the image to be detected, provided by an embodiment of the present application;

[0023] Figure 5 It is a schematic diagram of a process for extracting spatial features of light response non-uniform noise of an image to be detected and obtaining spatial feature information provided by an embodiment of the present application;

[0024] Figure 6 This is a flow chart of an embodiment of the present application, which provides a method of inputting spatial domain feature information and frequency domain feature information into an image tampering detection network to perform image tampering detection, and obtaining image detection information corresponding to an image to be detected;

[0025] Figure 7 is a schematic diagram of an image tampering detection network provided in an embodiment of the present application;

[0026] Figure 8 is a structural diagram of an image tampering detection network provided in an embodiment of the present application;

[0027] Fig. 9 It is a flowchart of a network training method provided in an embodiment of the present application;

[0028] Fig.10 It is a block diagram of a device for detecting image tampering provided by an embodiment of the present application;

[0029] Fig.11 It is a structural schematic diagram of an image tampering detection device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0031] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] It is understandable that in the specific implementation of this application, related data such as user information is involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of relevant countries and regions.

[0033] See also Figure 1 , Figure 1 It is a schematic diagram of an application environment provided by an embodiment of the present application, and the application environment may include a client 10 and a server 20, and the client 10 and the server 20 may be directly or indirectly connected via wired or wireless communication. The user may send an image tampering detection request to the server 20 via the client 10. The server 20 determines the corresponding image to be detected based on the image tampering detection request, and then performs noise feature recognition on the image to be detected to obtain spatial domain feature information of the light response non-uniform noise in the image to be detected and frequency domain feature information of the light response non-uniform noise, and then inputs the spatial domain feature information and the frequency domain feature information into the image tampering detection network to perform image tampering detection, obtain image detection information corresponding to the image to be detected, and return the image detection information to the client 10. It should be noted that Figure 1 Just an example.

[0034] The client can be a physical device such as a smartphone, a computer (such as a desktop computer, a tablet computer, a laptop computer), a digital assistant, an intelligent voice interaction device (such as an intelligent speaker), a smart wearable device, etc. It can also be software running in a physical device, such as a computer program. The operating system corresponding to the client can be Android system (Android system), iOS system (a mobile operating system developed by Apple), Linux system (an operating system), Microsoft Windows system (Microsoft Windows operating system), etc.

[0035] The server side can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The server can include a network communication unit, a processor, and a memory, etc. The server side can provide background services for the corresponding client.

[0036] The client 10 and the server 20 can be used to build a system for image tampering detection, which can be a distributed system. Taking the distributed system as a blockchain system as an example, it is formed by multiple nodes (any form of computing devices in the access network, such as servers, user terminals) and clients, and the nodes form a peer-to-peer (P2P, PeerTo Peer) network. The P2P protocol is an application layer protocol running on the Transmission Control Protocol (TCP, Transmission Control Protocol) protocol. In a distributed system, any machine such as a server or a terminal can join and become a node, and the node includes a hardware layer, an intermediate layer, an operating system layer, and an application layer.

[0037] The functions of each node in the above blockchain system include:

[0038] 1) Routing: a basic function of a node, used to support communication between nodes.

[0039] In addition to the routing function, the node can also have the following functions:

[0040] 2) Applications are deployed in the blockchain to implement specific businesses based on actual business needs, record data related to the implementation of functions to form record data, carry digital signatures in the record data to indicate the source of the task data, and send the record data to other nodes in the blockchain system for other nodes to add the record data to a temporary block when they successfully verify the source and integrity of the record data.

[0041] 3) Blockchain, including a series of blocks that are connected to each other in the order of their generation. Once a new block is added to the blockchain, it will not be removed. The block records the record data submitted by the nodes in the blockchain system.

[0042] The following describes a specific embodiment of an image tampering detection method provided by the present application. Figure 2 It is a flowchart of an image tampering detection method provided in an embodiment of the present application. The present application provides method operation steps as described in the embodiment or flowchart, but may include more or fewer operation steps based on conventional or non-creative labor. The order of steps listed in the embodiment is only one way of executing the steps among many orders, and does not represent the only order of execution. When the actual system or product is executed, it can be executed in the order of the method shown in the embodiment or the drawings or in parallel (for example, in a parallel processor or multi-threaded processing environment). Specifically, Figure 2 As shown, the method may include:

[0043] S201, obtaining an image to be detected.

[0044] In the embodiments of the present specification, the above-mentioned image to be detected may be any image that needs to be tampered with.

[0045] In practical applications, the size data of images from different sources may be different. In order to ensure the accuracy of the image tampering detection results, the initial image to be detected is usually preprocessed to obtain an image to be detected of a preset size. Specifically, the preset size can be set in combination with the accuracy requirements of image tampering detection in practical applications. In order to avoid the destruction of image pixel features that may be caused by scaling the image, in the embodiments of this specification, the image to be detected can be obtained by cropping the initial image to be detected.

[0046] In some embodiments, Figure 3 As shown, Figure 3 This is a schematic diagram of a process for obtaining an image to be detected provided by an embodiment of this specification, which may specifically include:

[0047] S301, obtaining an initial image to be detected.

[0048] Specifically, the size data of the initial image to be detected is generally not less than the above preset size.

[0049] S302, performing center cropping processing on the initial image to be detected to obtain the image to be detected.

[0050] In a specific embodiment, the center cropping process may be to perform center cropping on different channel data of the image according to the image channel data. The specific cropping area may include:

[0051] startXpos=(srcWidth–dstWidth) / 2;

[0052] startYpos=(srcHeight–dstHeight) / 2;

[0053] endXpos=startXpos+dstWidth;

[0054] endYpos=startYpos+dstHeight;

[0055] Among them, (srcWidth, srcHeight) is the width and height of the initial image to be detected, (dstWidth, dstHeight) is the width and height of the image to be detected, (startXpos, endXpos) is the starting position and ending position of the cropping area in the horizontal direction, and (startYpos, endYpos) is the starting position and ending position of the cropping area in the vertical direction.

[0056] It can be seen from the above embodiments that performing center cropping on the initial image to be detected to obtain the image to be detected can ensure the consistency of the size of the detected image, avoid the destruction of image pixel features that may be caused by image processing methods such as scaling, and improve the accuracy of noise feature extraction, thereby improving the accuracy of tampering detection results.

[0057] S202, performing noise feature recognition on the image to be detected to obtain spatial domain feature information and frequency domain feature information of the light response non-uniform noise in the image to be detected.

[0058] In practical applications, light response non-uniform noise is a kind of noise interference introduced into digital images due to hardware manufacturing defects of camera sensor CCD (Charge Coupled Device). Currently, CCDs manufactured in the industry have process problems, resulting in inconsistent thickness of semiconductor silicon materials for each pixel of CCD. When the camera is sensitive to light, slight differences will occur between different pixels. Such slight differences are brought into each pixel of the digital image to form a relatively weak noise. Therefore, in the embodiments of this specification, by performing feature recognition of light response non-uniform noise on the image to be detected, spatial feature information and frequency feature information of light response non-uniform noise are obtained, thereby performing image tampering detection on the image to be detected based on the spatial feature information and frequency feature information.

[0059] Specifically, the spatial domain feature information can characterize the spatial domain features of the light response non-uniform noise, and the frequency domain feature information can characterize the frequency domain features of the light response non-uniform noise.

[0060] In the embodiment of this specification, as shown in 4, the noise feature recognition is performed on the image to be detected to obtain the spatial domain feature information and the frequency domain feature information of the light response non-uniform noise in the image to be detected, which may include:

[0061] S401, extracting spatial features of light response non-uniform noise from an image to be detected to obtain spatial feature information.

[0062] In the embodiment of the present specification, the spatial feature information may be expressed in the form of a spatial feature image.

[0063] In a specific embodiment, Figure 5 As shown, the spatial feature extraction of the light response non-uniform noise of the image to be detected can obtain the spatial feature information, which may include:

[0064] S501 , extracting high-frequency components from multiple color channels of the image to be detected respectively to obtain high-frequency component data of the multiple color channels.

[0065] Specifically, the high-frequency component data may include: a horizontal high-frequency component, a vertical high-frequency component, and a diagonal high-frequency component.

[0066] In a specific embodiment, extracting high-frequency components from multiple color channels of the image to be detected respectively to obtain high-frequency component data of multiple color channels may include: performing multi-scale wavelet transform processing on multiple color channels of the image to be detected respectively to obtain wavelet domain high-frequency component data of multiple scales corresponding to each color channel. Optionally, the multi-scale wavelet transform processing may be a 4-level wavelet transform, and the wavelet basis may be db4 (4th-order Dobesi wavelet).

[0067] S502, performing noise intensity analysis on the high frequency component data of multiple color channels respectively to obtain noise intensity data.

[0068] Specifically, local noise variance analysis is performed on the horizontal high-frequency component, vertical high-frequency component and diagonal high-frequency component of each color channel to obtain target variance data of the horizontal high-frequency component, target variance data of the vertical high-frequency component and target variance data of the diagonal high-frequency component; the target variance data of the horizontal high-frequency component, the target variance data of the vertical high-frequency component and the target variance data of the diagonal high-frequency component are used as the above-mentioned noise intensity data.

[0069] In an optional embodiment, the local noise variance analysis may include performing noise variance analysis of multiple window sizes on each high-frequency component data, obtaining variance data corresponding to the multiple window sizes, and selecting the minimum value of the variance data corresponding to the multiple window sizes as the target variance data for each high-frequency component data. Preferably, the multiple windows may include: 3, 5, 7, 9, and the initial standard deviation in the local noise variance analysis process may be 5.

[0070] S503: Based on the noise intensity data, filter the high frequency component data of the multiple color channels respectively to obtain initial noise information of the multiple color channels.

[0071] In a specific embodiment, when the above-mentioned noise intensity data is the target variance data of the high-frequency component of each color channel, the above-mentioned filtering processing of the high-frequency component data of multiple color channels based on the noise intensity data to obtain the initial noise information of multiple color channels may include: using the target variance data of the high-frequency component of each color channel to filter the high-frequency component of each color channel to obtain the initial noise information of each color channel.

[0072] Optionally, the filter used in the filtering process may include but is not limited to: Wiener filter, Kalman filter, etc.

[0073] In a specific embodiment, when the high-frequency component is extracted by multi-scale wavelet transform processing, the initial noise information

[0074] S504, reconstructing the initial noise information of multiple color channels to obtain initial spatial domain feature information.

[0075] In a specific embodiment, when the high-frequency component is extracted as a multi-scale wavelet transform, the initial noise information may be wavelet domain noise information. Accordingly, reconstructing the initial noise information of multiple color channels to obtain initial spatial feature information may include: performing inverse wavelet transform on the wavelet domain noise information of multiple color channels to obtain spatial noise information of the image to be detected, and using the spatial noise information as the initial spatial feature information of the light response non-uniformity noise.

[0076] S505, performing noise enhancement processing on the initial spatial feature information to obtain spatial feature information.

[0077] In a specific embodiment, performing noise enhancement processing on the initial spatial feature information to obtain the spatial feature information may include:

[0078] 1) Perform zero-mean filtering on the initial spatial feature information to obtain the initial spatial feature information after filtering.

[0079] Specifically, interference information in the initial spatial feature information is removed by zero-mean filtering to enhance the light response non-uniformity noise.

[0080] 2) Perform Fourier peak suppression processing on the initial spatial feature information after filtering to obtain spatial feature information.

[0081] In practical applications, Fourier peak suppression processing can be used to remove the more glaring interference pixels in the image from the dimension of the Fourier frequency domain to enhance the light response non-uniform noise.

[0082] It can be seen from the above embodiments that by filtering the high-frequency component data of the image to be detected to obtain initial noise information, reconstructing the initial noise information to obtain initial spatial feature information, and then performing noise enhancement processing on the initial spatial feature information to obtain spatial feature information, interference can be effectively removed and the accuracy of spatial feature extraction can be improved.

[0083] S402, performing space-frequency transformation processing on the spatial domain feature information to obtain frequency domain feature information.

[0084] In the embodiment of the present specification, the frequency domain feature information may be expressed in the form of a frequency domain feature image.

[0085] In a specific embodiment, performing space-frequency transformation on the spatial domain feature information to obtain frequency domain feature information may include:

[0086] 1) Perform discrete Fourier transform on the spatial domain feature information to obtain the initial frequency domain feature information.

[0087] 2) Perform spectrum centering processing on the initial frequency domain feature information to obtain frequency domain feature information.

[0088] Specifically, when the initial frequency domain feature information is an initial frequency domain feature image, the DC component in the initial frequency domain feature image is usually located in the four vertex areas of the image, and the AC component in the initial frequency domain feature image is usually located in the remaining area of ​​the image. In order to facilitate subsequent feature extraction in the embodiment of the present specification, the initial frequency domain feature information can be spectrally centered to convert the DC component to the center of the image.

[0089] In an optional embodiment, the initial frequency domain feature image is divided based on the horizontal symmetry axis and the vertical symmetry axis of the image to obtain four regions, and the two regions at the diagonal corners of the image are interchanged respectively, so as to convert the DC component to the center of the image.

[0090] It can be seen from the above embodiments that performing discrete Fourier transform processing on the spatial domain feature information to obtain initial frequency domain feature information, and performing spectrum centering processing on the initial frequency domain feature information to obtain frequency domain feature information can effectively improve the accuracy of frequency domain feature extraction, thereby improving the accuracy of the detection results.

[0091] S203, inputting the spatial domain feature information and the frequency domain feature information into an image tampering detection network to perform image tampering detection, and obtaining image detection information corresponding to the image to be detected.

[0092] In an embodiment of the present specification, the above-mentioned image tampering detection network can be obtained by training a preset image tampering detection network for image tampering detection based on sample spatial domain feature information and sample frequency domain feature information. Specifically, the image tampering detection network may include: a spatial domain feature aggregation layer, a frequency domain feature aggregation layer, a feature fusion layer and a tampering detection layer.

[0093] In a specific embodiment, Figure 6 As shown, the spatial domain feature information and the frequency domain feature information are input into the image tampering detection network to perform image tampering detection, and the image detection information corresponding to the image to be detected may include:

[0094] S601, input the spatial feature information into the spatial feature aggregation layer for spatial feature aggregation processing to obtain target spatial feature information.

[0095] Specifically, the target spatial domain feature information can represent the spatial domain aggregation features of the light response non-uniform noise. The target spatial domain feature information can be expressed as a target spatial domain feature image.

[0096] In an optional embodiment, the spatial feature aggregation layer may include: a first convolution layer, a first batch of normalization layers, a first activation layer and at least one MBConv (Mobile Inverted Bottleneck Convolution) layer.

[0097] S602, inputting the frequency domain feature information into a frequency domain feature aggregation layer for frequency domain feature aggregation processing to obtain target frequency domain feature information.

[0098] Specifically, the target frequency domain feature information may represent the frequency domain aggregation feature of the light response non-uniform noise. The target frequency domain feature information may be expressed as a target frequency domain feature image.

[0099] In an optional embodiment, the frequency domain feature aggregation layer may include: a second convolution layer, a second batch normalization layer, a second activation layer and at least one MBConv (Mobile Inverted Bottleneck Convolution) layer.

[0100] In practical applications, the structures of the spatial feature aggregation layer and the frequency domain feature aggregation layer can be the same or different, but the sizes of the target spatial feature image output by the spatial feature aggregation layer and the target frequency domain feature image output by the frequency domain feature aggregation layer should be kept consistent to facilitate subsequent feature fusion.

[0101] S603, inputting the target spatial domain feature information and the target frequency domain feature information into a feature fusion layer for feature fusion processing to obtain target feature fusion information.

[0102] Specifically, the target frequency domain feature information can characterize the spatial domain aggregation features and frequency domain aggregation features of the light response non-uniformity noise.

[0103] In an optional embodiment, the feature fusion layer may be a channel fusion layer, and the target spatial domain feature information and the target frequency domain feature information are input into the feature fusion layer for feature fusion processing to obtain the target feature fusion information, which may include: inputting the target spatial domain feature information and the target frequency domain feature information into the channel fusion layer, superimposing and fusing the target spatial domain feature information and the target frequency domain feature information according to the image channel, and obtaining the target feature fusion information.

[0104] Specifically, the channel fusion layer can superimpose and fuse the target spatial feature information and the target frequency feature information from the dimension of the image channel. For example, the target spatial feature information is a spatial feature map with a size of 7×7×512, and the target frequency feature information is a frequency feature map with a size of 7×7×512, where 7×7 is the spatial resolution and 512 is the number of channels; accordingly, superimposing and fusing the target spatial feature information and the target frequency feature information from the dimension of the image channel may include: superimposing the spatial feature map and the frequency feature map according to the image channel to obtain a fused feature map of 7×7×1024, and using the fused feature map as the target feature fusion information.

[0105] In another optional embodiment, the feature fusion layer may include: a feature vector generation layer and a feature vector fusion layer. The above-mentioned inputting the target spatial domain feature information and the target frequency domain feature information into the feature fusion layer for feature fusion processing to obtain the target feature fusion information may include: inputting the target spatial domain feature information into the feature vector generation layer for spatial domain feature vector generation processing to obtain a spatial domain feature vector; inputting the target frequency domain feature information into the feature vector generation layer for frequency domain feature vector generation processing to obtain a frequency domain feature vector; inputting the spatial domain feature vector and the frequency domain feature vector into the feature vector fusion layer for feature vector fusion to obtain a target feature vector; and using the target feature vector as the target feature fusion information.

[0106] Specifically, the feature vector generation layer can be a global average pooling layer or a fully connected layer. Through the feature vector generation layer, the target spatial domain feature information and the target frequency domain feature information can be compressed respectively to obtain a spatial domain feature vector and a frequency domain feature vector. Through the feature vector fusion layer, the spatial domain feature vector and the frequency domain feature vector can be point-by-point added or multiplied from the dimension of the feature vector to obtain the target feature vector.

[0107] It can be seen from the above embodiments that a variety of feature fusion methods are provided. For different types of images to be detected, a feature fusion method of image channel fusion or fusion of image corresponding feature vectors can be selected to improve the accuracy of the target feature fusion information in representing the noise characteristics of the image to be detected.

[0108] S604, inputting the target feature fusion information into the tampering detection layer to perform image tampering detection to obtain image detection information.

[0109] In the embodiments of this specification, the image detection information can be used to indicate whether the image to be detected has been tampered with. Specifically, the image detection information can be a detection tag corresponding to the image to be detected, wherein the detection tag can be a real image tag or a tampered detection tag.

[0110] In a specific embodiment, the tampering detection layer may include a third convolutional layer, a global average pooling layer, a fully connected layer, and an output layer.

[0111] Specifically, the third convolution layer can perform convolution processing on the input target feature fusion information to achieve feature extraction of the target feature fusion information.

[0112] Specifically, the global average pooling layer can downsample the output of the previous layer, that is, return the maximum value in the sampling window as the downsampled output. On the one hand, it can make the image smaller and simplify the computational complexity; on the other hand, it can compress features and extract the main features.

[0113] Specifically, the fully connected layer can be used as a connection layer between the nodes of the upper and lower layers to establish a connection relationship between the node data obtained from the upper and lower layers. The fully connected layer can perform feature compression processing on the target feature fusion information to obtain the feature information to be detected.

[0114] Specifically, the classification layer can perform image tampering detection on the feature information to be detected and output the corresponding detection label. In a specific embodiment, the classification layer can use an activation function to output the target knowledge point label. In an optional embodiment, the activation function can be a Softmax function, which contains a nonlinear classifier for performing image tampering detection on the feature information to be detected.

[0115] In addition, it should be noted that the image tampering detection network described in the embodiments of this specification is not limited to the above-mentioned preset image tampering detection network. In practical applications, it may also include other machine learning networks, such as decision tree machine learning networks, etc. The embodiments of the present application are not limited to the above-mentioned preset image tampering detection network.

[0116] In a specific embodiment, Figure 7 As shown, an image tampering detection network including the spatial feature aggregation layer, the frequency domain feature aggregation layer, the feature fusion layer and the tampering detection layer is established, and the spatial feature information and the frequency domain feature information of the optical response non-uniform noise in the image to be detected are input into the image tampering detection network to perform image tampering detection, so as to obtain image detection information corresponding to the image to be detected.

[0117] In a specific embodiment, Figure 8 As shown, Figure 8 It is a structural diagram of an image tampering detection network provided by an embodiment of the present application. Specifically, the image tampering detection network may include: a spatial feature aggregation layer, a frequency domain feature aggregation layer, a feature fusion layer and a tampering detection layer; the spatial feature aggregation layer may include: a first convolution layer, a first batch of normalization layers, a first activation layer and 5 MBConv layers; the frequency domain feature aggregation layer may include: a second convolution layer, a second batch of normalization layers, a second activation layer and 5 MBConv layers; the tampering detection layer may include: a third convolution layer, a global average pooling layer, a fully connected layer and a classification layer.

[0118] In addition, it should be noted that the hierarchical structure of the spatial feature aggregation layer and the frequency domain feature aggregation layer in the embodiment of the present application is not fixed and can be designed arbitrarily, but the size of the target spatial feature image output by the spatial feature aggregation layer and the target frequency domain feature image output by the frequency domain feature aggregation layer should be kept consistent to facilitate subsequent feature fusion.

[0119] It can be seen from the above embodiments that the use of spatial feature aggregation layer, frequency domain feature aggregation layer, feature fusion layer and tampering detection layer for image tampering detection can improve the adaptability of tampering detection for different images, thereby greatly improving the accuracy of image tampering detection.

[0120] In the embodiments of the present specification, a preset image tampering detection network may be trained using sample detection images to obtain the above-mentioned image tampering detection network.

[0121] In a specific embodiment, Fig. 9 As shown, Fig. 9 This is a flowchart of a network training method provided by an embodiment of the present application, which may specifically include:

[0122] S901, obtaining a sample detection image and preset image detection information corresponding to the sample detection image.

[0123] In practical applications, before performing network training, training data may be determined first. Specifically, in an embodiment of the present application, a sample detection image containing preset image detection information may be obtained as training data.

[0124] Specifically, the preset image detection information may be a preset detection label pre-marked for the sample detection image. In an embodiment of the present specification, the preset detection label may include a real image label or a tampered image label, and the sample detection image may include a sample real image and a sample tampered image. Accordingly, the preset detection label of the sample real image may be a real image label, and the preset detection label of the sample tampered image may be a tampered image label. Generally, the ratio of the sample real image and the sample tampered image in the training data may be preset in combination with the actual requirements of the network training accuracy.

[0125] In a specific embodiment, the acquiring of the sample detection image may include:

[0126] 1) Obtain the initial sample detection image.

[0127] 2) Perform center cropping on the initial sample detection image to obtain a sample detection image.

[0128] In practical applications, by cropping the initial sample detection image for preprocessing, it is possible to ensure that the image data subsequently input into the preset image tampering detection network has the same size while avoiding destroying the characteristics of the light response non-uniform noise between image pixels. In addition, it can further reduce the training prediction time.

[0129] S902 , performing noise feature recognition on the sample detection image to obtain sample spatial domain feature information of the sample optical response non-uniform noise and sample frequency domain feature information of the sample optical response non-uniform noise in the sample detection image.

[0130] S903, inputting the sample spatial domain feature information and the sample frequency domain feature information into a preset image tampering detection network to perform image tampering detection, and obtaining sample image detection information corresponding to the sample detection image.

[0131] S904: Determine target loss information based on preset image detection information and sample image detection information.

[0132] S905, based on the target loss information, training a preset image tampering detection network to obtain an image tampering detection network.

[0133] In an optional embodiment, the sample image detection information may include a sample detection label of the sample detection image, and correspondingly, the target loss information may include a detection label loss;

[0134] Accordingly, the above-mentioned determination of target loss information based on the preset image detection information and the sample image detection information may include:

[0135] Determine the detection label loss based on the preset detection label and the sample detection label.

[0136] In a specific embodiment, determining the detection tag loss according to the preset detection tag and the sample detection tag may include determining the detection tag loss between the preset detection tag and the sample detection tag based on a preset loss function.

[0137] In a specific embodiment, the detection label loss may represent the difference between the preset detection label and the sample detection label.

[0138] In a specific embodiment, the preset loss function may include but is not limited to a cross entropy loss function, a logical loss function, an exponential loss function, etc.

[0139] In an optional embodiment, based on the target loss information, training a preset image tampering detection network to obtain the image tampering detection network may include:

[0140] S9051, based on the target loss information, updating the network parameters of the preset image tampering detection network;

[0141] S9052, based on the updated preset image tampering detection network, repeatedly performing the image tampering detection training iteration operation including steps S903, S904 and S9051 until the image tampering detection convergence condition is reached;

[0142] S9053, using the preset image tampering detection network obtained when the image tampering detection convergence condition is met as the image tampering detection network.

[0143] In an optional embodiment, the above-mentioned image tampering detection convergence condition can be that the number of training iteration operations reaches a preset number of training times. Optionally, the image tampering detection convergence condition can also be that the target loss information is less than a specified threshold. In the embodiments of this specification, the preset number of training times and the specified threshold can be pre-set in combination with the training speed and accuracy of the network in actual applications.

[0144] In a specific embodiment, the preset image tampering detection network may include a preset spatial feature aggregation layer, a preset frequency domain feature aggregation layer, a preset feature fusion layer and a preset tampering detection layer. Accordingly, the sample spatial feature information and the sample frequency domain feature information are input into the preset image tampering detection network for image tampering detection, and the sample image detection information corresponding to the sample detection image may include:

[0145] The sample spatial domain feature information is input into the preset spatial domain feature aggregation layer for spatial domain feature aggregation processing to obtain the sample target spatial domain feature information; the sample frequency domain feature information is input into the preset frequency domain feature aggregation layer for frequency domain feature aggregation processing to obtain the sample target frequency domain feature information; the sample target spatial domain feature information and the sample target frequency domain feature information are input into the preset feature fusion layer for feature fusion processing to obtain sample target feature fusion information; the sample target feature fusion information is input into the preset tampering detection layer for image tampering detection to obtain sample image detection information.

[0146] It can be seen from the above embodiments that, on the one hand, machine learning training based on sample detection images and corresponding preset detection labels can improve the generalization ability and robustness of the image tampering detection network, thereby better improving the accuracy of the network in detecting image tampering.

[0147] It can be seen from the technical solutions provided by the above embodiments of the present application that, in the scenario of image tampering detection, the present application, on the one hand, performs center cropping processing on the initial image to be detected to obtain the image to be detected, which can ensure the consistency of the size of the detected image, avoid the destruction of the image pixel features that may be caused by image processing methods such as scaling processing, and improve the accuracy of noise feature recognition; on the other hand, by filtering the high-frequency component data of the image to be detected to obtain initial noise information, reconstructing the initial noise information to obtain initial spatial feature information, and then performing noise enhancement processing on the initial spatial feature information to obtain spatial feature information, it can effectively remove interference and improve the accuracy of spatial feature extraction; on the other hand, On the one hand, the spatial domain feature information is processed by discrete Fourier transform to obtain the initial frequency domain feature information, and the initial frequency domain feature information is processed by spectrum centering to obtain frequency domain feature information, which can effectively improve the accuracy of frequency domain feature extraction; on the other hand, the spatial feature aggregation layer, frequency domain feature aggregation layer, feature fusion layer and tampering detection layer are used to detect image tampering, which can improve the adaptability of tampering detection of different images, and thus greatly improve the accuracy of image tampering detection; on the other hand, machine learning training based on sample detection images and corresponding preset detection labels can improve the generalization ability and robustness of the image tampering detection network, thereby better improving the accuracy of the network in image tampering detection.

[0148] The present application also provides an image tampering detection device. Fig.10 As shown, the image tampering detection device may include:

[0149] The image acquisition module 1010 is used to acquire the image to be detected;

[0150] The noise feature recognition module 1020 is used to perform noise feature recognition on the image to be detected, and obtain spatial domain feature information and frequency domain feature information of the light response non-uniform noise in the image to be detected;

[0151] The image tampering detection module 1030 is used to input the spatial domain feature information and the frequency domain feature information into the image tampering detection network to perform image tampering detection and obtain image detection information corresponding to the image to be detected.

[0152] In some embodiments, the above-mentioned to-be-detected image acquisition module 1010 may include:

[0153] An initial image to be detected acquiring unit, used for acquiring an initial image to be detected;

[0154] The center cropping processing unit is used to perform center cropping processing on the initial image to be detected to obtain the image to be detected.

[0155] In the embodiment of this specification, the noise feature recognition module 1020 may include:

[0156] A spatial feature extraction unit is used to extract the spatial features of the light response non-uniform noise of the image to be detected to obtain spatial feature information;

[0157] The space-frequency transform processing unit is used to perform space-frequency transform processing on the spatial domain feature information to obtain frequency domain feature information.

[0158] In a specific embodiment, the above-mentioned spatial feature extraction unit may include:

[0159] A high-frequency component extraction unit is used to extract high-frequency components of multiple color channels of the image to be detected, respectively, to obtain high-frequency component data of the multiple color channels;

[0160] A noise intensity analysis unit, used to perform noise intensity analysis on the high frequency component data of multiple color channels respectively to obtain noise intensity data;

[0161] A filtering processing unit, used to filter the high-frequency component data of multiple color channels respectively based on the noise intensity data to obtain initial noise information of the multiple color channels;

[0162] A reconstruction processing unit, used to reconstruct the initial noise information of multiple color channels to obtain initial spatial domain feature information;

[0163] The noise enhancement processing unit is used to perform noise enhancement processing on the initial spatial domain feature information to obtain the spatial domain feature information.

[0164] In a specific embodiment, the image tampering detection module 1030 may include:

[0165] A spatial feature aggregation processing unit is used to input the spatial feature information into the spatial feature aggregation layer for spatial feature aggregation processing to obtain target spatial feature information;

[0166] A frequency domain feature aggregation processing unit is used to input the frequency domain feature information into the frequency domain feature aggregation layer for frequency domain feature aggregation processing to obtain target frequency domain feature information;

[0167] A feature fusion unit is used to input the target spatial domain feature information and the target frequency domain feature information into the feature fusion layer for feature fusion processing to obtain target feature fusion information;

[0168] The image tampering detection unit is used to input the target feature fusion information into the tampering detection layer to perform image tampering detection and obtain image detection information.

[0169] In an optional embodiment, the feature fusion unit may include:

[0170] The stacking fusion unit is used to input the target spatial domain feature information and the target frequency domain feature information into the channel fusion layer, stack and fuse the target spatial domain feature information and the target frequency domain feature information according to the image channel to obtain the target feature fusion information.

[0171] In a specific embodiment, the above device may further include:

[0172] A sample acquisition module is used to acquire a sample detection image and preset image detection information corresponding to the sample detection image;

[0173] A sample noise feature recognition module is used to perform noise feature recognition on a sample detection image to obtain sample spatial domain feature information of sample light response non-uniform noise and sample frequency domain feature information of sample light response non-uniform noise in the sample detection image;

[0174] A sample image detection information module is used to input sample spatial domain feature information and sample frequency domain feature information into a preset image tampering detection network to perform image tampering detection, and obtain sample image detection information corresponding to the sample detection image;

[0175] A target loss information determination module, used to determine target loss information based on preset image detection information and sample image detection information;

[0176] The network training module is used to train a preset image tampering detection network based on target loss information to obtain an image tampering detection network.

[0177] It should be noted that the device and method embodiments in the device embodiment are based on the same inventive concept.

[0178] An embodiment of the present application provides an image tampering detection device, which includes a processor and a memory, in which at least one instruction or at least one program is stored, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the image tampering detection method provided in the above method embodiment.

[0179] Further, Fig.11 The hardware structure diagram of an image tampering detection device for implementing the image tampering detection method provided in the embodiment of the present application is shown. The image tampering detection device may participate in or include the image tampering detection device provided in the embodiment of the present application. Fig.11 As shown, the image tampering detection device 110 may include one or more (1102a, 1102b, ..., 1102n are used to illustrate) processors 1102 (the processor 1102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 1104 for storing data, and a transmission device 1106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Fig.11 The structure shown is only for illustration and does not limit the structure of the above electronic device. Fig.11 More or fewer components as shown, or with Fig.11 Different configurations are shown.

[0180] It should be noted that the one or more processors 1102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuit may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be fully or partially integrated into any of the other components in the image tampering detection device 110 (or mobile device). As involved in the embodiments of the present application, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0181] The memory 1104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the image tampering detection method described in the embodiment of the present application. The processor 1102 executes various functional applications and data processing by running the software programs and modules stored in the memory 1104, that is, to implement the above-mentioned image tampering detection method. The memory 1104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 1104 may further include a memory remotely arranged relative to the processor 1102, and these remote memories may be connected to the image tampering detection device 110 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.

[0182] The transmission device 1106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the image tampering detection device 110. In one example, the transmission device 1106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one embodiment, the transmission device 1106 can be a radio frequency (Radio Frequency, RF) module, which is used to communicate with the Internet wirelessly.

[0183] The display may be, for example, a touch screen liquid crystal display (LCD) that may enable a user to interact with a user interface of the image tampering detection device 110 (or mobile device).

[0184] An embodiment of the present application also provides a computer-readable storage medium, which can be set in an image tampering detection device to store at least one instruction or at least one program related to the image tampering detection method in the method embodiment. The at least one instruction or the at least one program is loaded and executed by the processor to implement the image tampering detection method provided by the above-mentioned method embodiment.

[0185] Optionally, in this embodiment, the storage medium may be located in at least one of the multiple network servers of the computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0186] The embodiment of the present application also provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the image tampering detection method provided in the method embodiment.

[0187] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of the present application are described. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0188] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and equipment embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0189] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0190] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for detecting image tampering, characterized in that: The method comprises: Acquire the image to be detected; Performing noise feature recognition on the image to be detected to obtain spatial domain feature information and frequency domain feature information of the light response non-uniform noise in the image to be detected, wherein the spatial domain feature information represents the spatial domain feature of the light response non-uniform noise, and the frequency domain feature information represents the frequency domain feature of the light response non-uniform noise; Inputting the spatial feature information into a spatial feature aggregation layer in an image tampering detection network for spatial feature aggregation processing to obtain target spatial feature information, wherein the target spatial feature information represents spatial domain aggregation features of the light response non-uniformity noise, and the spatial feature aggregation layer includes: a first convolution layer, a first batch of normalization layers, a first activation layer, and at least one moving flip bottleneck convolution layer; Inputting the frequency domain feature information into the frequency domain feature aggregation layer in the image tampering detection network for frequency domain feature aggregation processing to obtain target frequency domain feature information, wherein the target frequency domain feature information represents the frequency domain aggregation features of the light response non-uniform noise, wherein the frequency domain feature aggregation layer comprises: a second convolution layer, a second batch normalization layer, a second activation layer, and at least one layer of moving flip bottleneck convolution layer, wherein the spatial domain feature aggregation layer and the frequency domain feature aggregation layer adopt the same network structure, and the size of the target spatial domain feature information and the target frequency domain feature information are kept consistent; Input the target spatial domain feature information and the target frequency domain feature information into the feature fusion layer in the image tampering detection network, select a feature fusion method for the type of the image to be detected from multiple feature fusion methods, perform feature fusion on the target spatial domain feature information and the target frequency domain feature information to obtain target feature fusion information, wherein the target feature fusion information is fused with the spatial domain aggregation feature and the frequency domain aggregation feature of the light response non-uniform noise, and the multiple feature fusion methods include: an image channel stacking fusion method and a feature vector point-by-point fusion method; Inputting the target feature fusion information into the tampering detection layer in the image tampering detection network to perform image tampering detection, and obtaining image detection information corresponding to the image to be detected; Among them, the image tampering detection network is obtained by training the preset spatial feature aggregation layer, the preset frequency domain feature aggregation layer, the preset feature fusion layer and the preset tampering detection layer for image tampering detection based on the sample real image and the sample tampered image; the ratio of the sample real image and the sample tampered image is preset based on the training accuracy requirement of the image tampering detection network.

2. The method according to claim 1, characterized in that The performing noise feature recognition on the image to be detected to obtain spatial domain feature information of the light response non-uniform noise in the image to be detected and frequency domain feature information of the light response non-uniform noise includes: Extracting the spatial features of the light response non-uniform noise from the image to be detected to obtain the spatial feature information; Performing space-frequency transform processing on the spatial domain feature information to obtain the frequency domain feature information.

3. The method according to claim 2, characterized in that The extracting the spatial feature of the light response non-uniform noise from the image to be detected to obtain the spatial feature information includes: Extracting high-frequency components from multiple color channels of the image to be detected respectively to obtain high-frequency component data of the multiple color channels; Performing noise intensity analysis on the high frequency component data of the multiple color channels respectively to obtain noise intensity data; Based on the noise intensity data, filtering the high frequency component data of the multiple color channels respectively to obtain initial noise information of the multiple color channels; Reconstructing the initial noise information of the multiple color channels to obtain initial spatial domain feature information; The initial spatial domain feature information is subjected to noise enhancement processing to obtain the spatial domain feature information.

4. The method according to any one of claims 1 to 3, characterized in that: The acquiring of the image to be detected comprises: Obtain an initial image to be detected; Performing center cropping processing on the initial image to be detected to obtain the image to be detected.

5. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Acquire a sample detection image and preset image detection information corresponding to the sample detection image; Performing noise feature recognition on the sample detection image to obtain sample spatial domain feature information of sample light response non-uniform noise in the sample detection image and sample frequency domain feature information of the sample light response non-uniform noise; Inputting the sample spatial domain feature information and the sample frequency domain feature information into a preset image tampering detection network to perform image tampering detection, and obtaining sample image detection information corresponding to the sample detection image; Determining target loss information based on the preset image detection information and the sample image detection information; Based on the target loss information, the preset image tampering detection network is trained to obtain the image tampering detection network.

6. The method according to claim 5, characterized in that The preset image tampering detection network includes a preset spatial feature aggregation layer, a preset frequency feature aggregation layer, a preset feature fusion layer and a preset tampering detection layer, and the sample spatial feature information and the sample frequency feature information are input into the preset image tampering detection network for image tampering detection, and the sample image detection information corresponding to the sample detection image is obtained, including: Inputting the sample spatial feature information into the preset spatial feature aggregation layer for spatial feature aggregation processing to obtain sample target spatial feature information; Inputting the sample frequency domain feature information into the preset frequency domain feature aggregation layer for frequency domain feature aggregation processing to obtain sample target frequency domain feature information; Inputting the sample target spatial domain feature information and the sample target frequency domain feature information into the preset feature fusion layer for feature fusion processing to obtain sample target feature fusion information; The sample target feature fusion information is input into the preset tampering detection layer to perform image tampering detection to obtain the sample image detection information.

7. An image tampering detection device, characterized in that: The device comprises: An image acquisition module to be detected is used to acquire the image to be detected; A noise feature recognition module, used to perform noise feature recognition on the image to be detected, and obtain spatial domain feature information and frequency domain feature information of the light response non-uniform noise in the image to be detected, wherein the spatial domain feature information represents the spatial domain feature of the light response non-uniform noise, and the frequency domain feature information represents the frequency domain feature of the light response non-uniform noise; A spatial feature aggregation processing module is used to input the spatial feature information into a spatial feature aggregation layer in an image tampering detection network for spatial feature aggregation processing to obtain target spatial feature information, wherein the target spatial feature information represents the spatial domain aggregation features of the light response non-uniform noise, and the spatial feature aggregation layer includes: a first convolution layer, a first batch of normalization layers, a first activation layer, and at least one moving flip bottleneck convolution layer; A frequency domain feature aggregation processing module, used for inputting the frequency domain feature information into the frequency domain feature aggregation layer in the image tampering detection network for frequency domain feature aggregation processing to obtain target frequency domain feature information, wherein the target frequency domain feature information represents the frequency domain aggregation features of the light response non-uniform noise, the frequency domain feature aggregation layer comprises: a second convolution layer, a second batch normalization layer, a second activation layer and at least one layer of moving flip bottleneck convolution layer, the spatial domain feature aggregation layer and the frequency domain feature aggregation layer adopt the same network structure, and the size of the target spatial domain feature information and the target frequency domain feature information are consistent; A feature fusion module is used to input the target spatial domain feature information and the target frequency domain feature information into a feature fusion layer in the image tampering detection network, select a feature fusion method for the type of the image to be detected from a plurality of feature fusion methods, perform feature fusion on the target spatial domain feature information and the target frequency domain feature information to obtain target feature fusion information, wherein the target feature fusion information is fused with the spatial domain aggregation feature and the frequency domain aggregation feature of the light response non-uniform noise, wherein the plurality of feature fusion methods include: an image channel stacking fusion method and a feature vector point-by-point fusion method; An image tampering detection module is used to input the target feature fusion information into the tampering detection layer in the image tampering detection network to perform image tampering detection, and obtain image detection information corresponding to the image to be detected; Among them, the image tampering detection network is obtained by training the preset spatial feature aggregation layer, the preset frequency domain feature aggregation layer, the preset feature fusion layer and the preset tampering detection layer for image tampering detection based on the sample real image and the sample tampered image; the ratio of the sample real image and the sample tampered image is preset based on the training accuracy requirement of the image tampering detection network.

8. The device according to claim 7, characterized in that The noise feature recognition module comprises: A spatial feature extraction unit, used for performing spatial feature extraction of the light response non-uniform noise on the image to be detected to obtain the spatial feature information; The space-frequency transform processing unit is used to perform space-frequency transform processing on the spatial domain feature information to obtain the frequency domain feature information.

9. The device according to claim 8, characterized in that The spatial feature extraction unit comprises: A high-frequency component extraction unit, used to extract high-frequency components from a plurality of color channels of the image to be detected, respectively, to obtain high-frequency component data of the plurality of color channels; A noise intensity analysis unit, configured to perform noise intensity analysis on the high frequency component data of the plurality of color channels respectively to obtain noise intensity data; A filtering processing unit, configured to perform filtering processing on the high frequency component data of the plurality of color channels respectively based on the noise intensity data, so as to obtain initial noise information of the plurality of color channels; A reconstruction processing unit, used to reconstruct the initial noise information of the multiple color channels to obtain initial spatial domain feature information; The noise enhancement processing unit is used to perform noise enhancement processing on the initial spatial domain feature information to obtain the spatial domain feature information.

10. The device according to any one of claims 7 to 9, characterized in that: The to-be-detected image acquisition module comprises: An initial image to be detected acquiring unit, used for acquiring an initial image to be detected; The center cropping processing unit is used to perform center cropping processing on the initial image to be detected to obtain the image to be detected.

11. The device according to any one of claims 7 to 9, characterized in that: The device also includes: A sample acquisition module, used to acquire a sample detection image and preset image detection information corresponding to the sample detection image; A sample noise feature recognition module, used to perform noise feature recognition on the sample detection image to obtain sample spatial domain feature information of the sample light response non-uniform noise in the sample detection image and sample frequency domain feature information of the sample light response non-uniform noise; A sample image detection information module is used to input the sample spatial domain feature information and the sample frequency domain feature information into a preset image tampering detection network to perform image tampering detection, and obtain sample image detection information corresponding to the sample detection image; A target loss information determination module, used to determine the target loss information based on the preset image detection information and the sample image detection information; A network training module is used to train the preset image tampering detection network based on the target loss information to obtain the image tampering detection network.

12. The device according to claim 11, characterized in that The preset image tampering detection network includes a preset spatial feature aggregation layer, a preset frequency domain feature aggregation layer, a preset feature fusion layer and a preset tampering detection layer, and the sample image detection information module is further used to: Inputting the sample spatial feature information into the preset spatial feature aggregation layer for spatial feature aggregation processing to obtain sample target spatial feature information; Inputting the sample frequency domain feature information into the preset frequency domain feature aggregation layer for frequency domain feature aggregation processing to obtain sample target frequency domain feature information; Inputting the sample target spatial domain feature information and the sample target frequency domain feature information into the preset feature fusion layer for feature fusion processing to obtain sample target feature fusion information; The sample target feature fusion information is input into the preset tampering detection layer to perform image tampering detection to obtain the sample image detection information.

13. An image tampering detection device, characterized in that: The device includes a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the image tampering detection method as described in any one of claims 1 to 6.

14. A computer-readable storage medium, characterized in that: The storage medium stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the image tampering detection method as described in any one of claims 1 to 6.

15. A computer program product, characterized in that The computer program product includes at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the image tampering detection method according to any one of claims 1 to 6.

Images