Hotel platform page screenshot forgery detection method, device, medium and program product

Abstract

This invention discloses a method, device, medium, and program product for detecting forgery in screenshots of hotel platform pages. The method includes: first, acquiring a screenshot of the hotel platform page to be detected; performing multi-dimensional forgery detection on the screenshot to obtain multi-dimensional detection results, including image tampering detection, screenshot source detection, and price area anomaly detection; finally, inputting the multi-dimensional detection results into a preset large language model for confidence and feature analysis processing to generate a forgery detection report. This invention effectively improves the efficiency and accuracy of forgery detection for hotel platform page screenshots.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to methods, equipment, media, and programs for detecting forgery in screenshots of hotel platform pages. Background Technology

[0002] In the operation of modern travel apps, the newly launched hotel price comparison function allows users to upload screenshots of hotel pages from other platforms and compare prices with the app's own, thus finding more affordable hotel options. This function aims to improve user experience and platform competitiveness. However, this function also faces the risk of malicious users manipulating hotel prices. These users may obtain lower-priced hotels through irregular means (such as altering price information in screenshots), causing economic losses and reputational damage to the platform.

[0003] Currently, existing technologies for detecting tampering of uploaded screenshots mainly rely on the following traditional methods: 1. Manual review: Platform staff manually check the authenticity of screenshots, relying on experience to make judgments, but this is inefficient and cannot handle massive uploads; 2. Simple OCR recognition: Only extracts text information for comparison, but cannot detect image-level tampering traces, such as splicing, noise inconsistencies, etc.; 3. Metadata inspection: Preliminary review of EXIF ​​data, but lacks comprehensive analysis and cannot deal with complex forgery methods.

[0004] The traditional methods described above have the following significant limitations: 1. Low detection accuracy: It is difficult to identify advanced tampering, such as inconsistent noise patterns or copy-and-move operations; 2. Response delay: Manual review leads to long processing times and cannot prevent fraudulent transactions in real time; 3. Lack of quantitative evaluation: Judgment relies heavily on subjective experience and lacks objective confidence scores and feature analysis; 4. Information silos exist: Different detection modules (such as image tampering and price area analysis) are not integrated, making it impossible to form a closed-loop decision-making mechanism.

[0005] Therefore, there is an urgent need to invent an automated, intelligent, and comprehensive method for detecting forged screenshots of hotel platform pages, in order to solve the problems of low efficiency and poor accuracy in existing forgery detection technologies. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a method, device, medium, and program product for detecting forgery of screenshots of hotel platform pages, which at least partially solves the problems existing in the prior art.

[0007] Other features and advantages of the invention will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0008] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0009] According to a first aspect of the present invention, a method for detecting forgery of screenshots of hotel platform pages is provided, the method comprising:

[0010] Obtain screenshots of the hotel platform pages to be tested;

[0011] Multi-dimensional forgery detection is performed on the screenshot to be detected to obtain multi-dimensional detection results. The multi-dimensional forgery detection includes image tampering detection, screenshot source detection, and price area anomaly detection.

[0012] The multi-dimensional detection results are input into a preset large language model for confidence and feature analysis to generate a forgery detection report.

[0013] Furthermore, multi-dimensional forgery detection is performed on the screenshot to be detected, resulting in multi-dimensional detection results, including:

[0014] The screenshot to be detected is subjected to image tampering detection to obtain the tampering detection result;

[0015] If the tampering detection result indicates that there are traces of tampering, then the screenshot to be detected will be marked as a high-risk screenshot;

[0016] If the tampering detection result is no trace of tampering, then the screenshot to be detected is subjected to a screenshot source authenticity detection to obtain the screenshot source detection result;

[0017] If the screenshot source detection result indicates that the source is not genuine, then the screenshot to be detected will be marked as a high-risk screenshot;

[0018] If the screenshot source detection result indicates that the source is genuine, then price area anomaly detection is performed on the screenshot to be detected to obtain the price anomaly detection result;

[0019] If the price anomaly detection result indicates the presence of a price anomaly, the screenshot to be detected will be marked as a high-risk screenshot.

[0020] If the price anomaly detection result is that there is no price anomaly, then the screenshot to be detected will not be marked as a high-risk screenshot.

[0021] Further, image tampering detection is performed on the screenshot to be detected to obtain tampering detection results, including:

[0022] Read the EXIF ​​metadata of the screenshot to be detected, and detect whether the EXIF ​​metadata contains image editing software tag information, wherein the image editing software tag information includes the image editing software name and / or image editing software version information;

[0023] If the EXIF ​​metadata contains image editing software tag information, the tampering detection result indicates that there are traces of tampering.

[0024] The screenshot to be detected is recompressed using JPEG to obtain a recompressed screenshot;

[0025] Calculate the pixel differences between the recompressed screenshot and the original screenshot to be detected, and generate a visualized ELA difference map;

[0026] Detect whether there are regions with inconsistent compression history in the ELA difference map;

[0027] If there are areas of inconsistent compression history in the ELA difference map, the tampering detection result indicates that there are traces of tampering.

[0028] A consistency analysis of the noise patterns in each region of the screenshot to be detected is performed to detect whether there are regions with abnormal noise patterns in the screenshot to be detected.

[0029] If there is an abnormal noise pattern area in the screenshot to be detected, the tampering detection result is considered to have been tampered with;

[0030] Feature matching is performed within the screenshot to be detected to detect whether there are two or more regions in the screenshot to be detected with a similarity greater than a preset similarity threshold.

[0031] If there are two or more regions in the screenshot to be detected that have a similarity greater than a preset similarity threshold, the tampering detection result is that there are traces of tampering.

[0032] If the EXIF ​​metadata does not contain image editing software tag information, and the ELA difference map does not contain areas with inconsistent compression history, and the screenshot to be detected does not contain areas with abnormal noise patterns, and the screenshot to be detected does not contain more than two areas with similarity greater than a preset similarity threshold, then the tampering detection result is no tampering trace.

[0033] Furthermore, the screenshot to be detected is subjected to a source authenticity detection to obtain the screenshot source detection result, including:

[0034] Perform a Fourier transform on the screenshot to be detected to obtain the frequency domain features of the screenshot;

[0035] Based on the frequency domain characteristics of the screenshot, determine whether there are traces of digital generation in the screenshot to be detected;

[0036] If the screenshot to be detected contains traces of digital generation, then the screenshot source detection result is that the source is not genuine;

[0037] If there are no traces of digital generation in the screenshot to be detected, then the screenshot source detection result is that the source is genuine.

[0038] Further, price region anomaly detection is performed on the screenshot to be detected to obtain price anomaly detection results, including:

[0039] By using ORC to identify and preset hotel price data templates, candidate price areas are located in the screenshot to be detected.

[0040] Non-maximum suppression is applied to the candidate price region to remove duplicate boxes, resulting in the final price region.

[0041] Image tampering detection is performed on the final price area to obtain the price area tampering detection result;

[0042] If there are more than two final price regions, then a content consistency check is performed on each final price region to obtain the price consistency check result.

[0043] Furthermore, the multi-dimensional detection results are input into a preset large language model for confidence and feature analysis to generate a forgery detection report, including:

[0044] The tampering detection result, the screenshot source detection result, the price anomaly detection result, and the high-risk screenshot are input as multi-dimensional detection results into a preset large language model;

[0045] The multi-dimensional detection results are subjected to semantic analysis and confidence quantification by the preset large language model to obtain semantic feature analysis results and confidence scores;

[0046] Based on the semantic feature analysis results and the confidence score, a forgery detection report is generated.

[0047] According to a second aspect of the present invention, a forgery detection device for screenshots of hotel platform pages is provided, the device comprising: a processor and a memory;

[0048] The memory is used to store one or more program instructions;

[0049] The processor is configured to run one or more program instructions to perform the steps of the forgery detection method for hotel platform page screenshots as described in any of the preceding claims.

[0050] According to a third aspect of the present invention, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and the computer program, when executed by a processor, implements the steps of the forgery detection method for hotel platform page screenshots as described in any of the preceding claims.

[0051] According to a fourth aspect of the present invention, a computer program product is provided, the computer program product comprising a computing program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions that, when executed by a computer, cause the computer to implement the steps of the forgery detection method for hotel platform page screenshots as described in any of the preceding claims.

[0052] This invention provides a method, device, medium, and program product for detecting forgery in hotel platform page screenshots. The method includes: first, acquiring a screenshot of the hotel platform page to be detected; performing multi-dimensional forgery detection on the screenshot to obtain multi-dimensional detection results, including image tampering detection, screenshot source detection, and price area anomaly detection; finally, inputting the multi-dimensional detection results into a preset large language model for confidence and feature analysis processing to generate a forgery detection report. This invention effectively improves the efficiency and accuracy of forgery detection for hotel platform page screenshots. Attached Figure Description

[0053] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0054] Figure 1 This is a flowchart illustrating a method for detecting forgery of screenshots from hotel platform pages provided in an embodiment of the present invention.

[0055] Figure 2 A schematic diagram of the architecture of a computer program product provided in an embodiment of the present invention;

[0056] Figure 3 This is a schematic diagram illustrating the principle of a method for detecting forgery of screenshots from hotel platform pages provided in an embodiment of the present invention.

[0057] Figure 4 This is a schematic diagram illustrating the content of a forged detection report provided in an embodiment of the present invention. Detailed Implementation

[0058] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0059] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0060] Figure 1 A flowchart of a method for detecting forgery of screenshots of hotel platform pages according to an embodiment of the present invention is shown.

[0061] like Figure 1 As shown, the method for detecting forgery of screenshots of hotel platform pages according to an embodiment of the present invention may include steps S100, S200 and S300.

[0062] In step S100, a screenshot of the hotel platform page to be tested is obtained.

[0063] In step S200, multi-dimensional forgery detection is performed on the screenshot to be detected to obtain multi-dimensional detection results.

[0064] Specifically, the above steps include:

[0065] The process involves several steps: First, image tampering detection is performed on the screenshot to be tested, and the tampering detection result is obtained. If the tampering detection result indicates the presence of tampering traces, the screenshot to be tested is marked as a high-risk screenshot. If the tampering detection result indicates no tampering traces, the screenshot to be tested is then subjected to a screenshot source authenticity detection, and the screenshot source detection result is obtained. If the screenshot source detection result indicates an authentic source, the screenshot to be tested is marked as a high-risk screenshot. If the screenshot source detection result indicates an authentic source, the screenshot to be tested is then subjected to a price anomaly detection, and the price anomaly detection result is obtained. If the price anomaly detection result indicates the presence of a price anomaly, the screenshot to be tested is marked as a high-risk screenshot. If the price anomaly detection result indicates no price anomaly, the screenshot to be tested is not marked as a high-risk screenshot.

[0066] More specifically, image tampering detection is performed on the screenshot to be detected, and the tampering detection results are obtained, including:

[0067] The EXIF ​​metadata of the screenshot to be detected is read, and it is checked whether the EXIF ​​metadata contains image editing software tag information. This tag information includes the image editing software name and / or version information. If the EXIF ​​metadata contains image editing software tag information, the tampering detection result indicates that tampering has occurred. The principle behind this EXIF ​​metadata analysis is that many image editing software programs (such as Adobe Photoshop) automatically write the software name and version information into the EXIF ​​data when saving images, thus providing direct evidence that the image has been tampered with.

[0068] The screenshot to be detected is recompressed using JPEG to obtain a recompressed screenshot. The pixel difference between the recompressed screenshot and the original screenshot to be detected is calculated to generate a visual ELA difference map. The ELA difference map is then checked for regions with inconsistent compression histories. If such regions exist, the tampering detection result indicates tampering. The principle behind this ELA difference analysis is as follows: JPEG is a lossy compression method; an unmodified image should have a uniform overall compression loss (error level); however, tampered and spliced ​​areas, due to their different compression histories compared to the background, will exhibit significantly different brightness (usually brighter) on the ELA analysis map.

[0069] A consistency analysis is performed on the noise patterns (such as noise intensity and distribution) of each region in the screenshot to be tested, detecting whether there are areas with abnormal noise patterns in the screenshot. If areas with abnormal noise patterns are found in the screenshot, the tampering detection result indicates that tampering has occurred. The principle of the above noise analysis is as follows: any digital camera sensor will introduce a weak and unique noise pattern (camera fingerprint) during imaging; and the noise pattern of the stitched area will be inconsistent with the noise pattern of the original background, thus exposing its external source.

[0070] Feature matching is performed within the screenshot to be detected to check if there are two or more regions in the screenshot with a similarity greater than a preset similarity threshold. If there are two or more regions in the screenshot with a similarity greater than the preset similarity threshold, the tampering detection result is that tampering has occurred. The principle of the copy-move tampering detection above is: in order to cover up or copy an object, the most common method used by tamperers is to copy content from one area of ​​the image and then move it to another area, resulting in unnatural repeating patterns appearing within the image.

[0071] If the EXIF ​​metadata does not contain image editing software tag information, and there are no inconsistent compression history areas in the ELA difference map, and there are no abnormal noise patterns in the screenshot to be detected, and there are no more than two areas with similarity greater than the preset similarity threshold in the screenshot to be detected, then the tampering detection result is no tampering trace.

[0072] More specifically, the screenshot to be detected undergoes a source authenticity check to obtain the screenshot source detection results, including:

[0073] Perform a Fourier transform on the screenshot to be detected to obtain the screenshot's frequency domain features. Based on the screenshot's frequency domain features, check pixel-level consistency, edge artifacts, and device fingerprints to determine whether there are traces of digital generation in the screenshot to be detected. If traces of digital generation are found in the screenshot to be detected, the screenshot source detection result is that the source is not genuine. If there are no traces of digital generation in the screenshot to be detected, the screenshot source detection result is that the source is genuine.

[0074] The principle behind the above screenshot authenticity detection is as follows: Real physical screenshots (such as mobile phone photos) introduce real lighting, reflections, and sensor noise, while digital generation (such as software simulation) often lacks these natural traces. This invention identifies whether the image was captured by a real device by comparing the frequency domain features (such as Fourier transform) of the image.

[0075] More specifically, price anomaly detection is performed on the screenshot to be detected, and the price anomaly detection results are obtained, including:

[0076] By using ORC recognition, pre-set hotel price data templates, and identifier templates such as ¥, candidate price regions are located in the screenshot to be detected. Non-maximum suppression is applied to the candidate price regions to remove duplicate boxes, resulting in the final price regions. Image tampering detection is performed on the final price regions to obtain the price region tampering detection results. If there are more than two final price regions, content consistency detection is performed on each final price region to obtain the price consistency detection results. For example, if some tamperers only tamper with a certain price on the hotel list page, this comparison will detect the price inconsistency.

[0077] Finally, in step S300, the multi-dimensional detection results are input into a preset large language model for confidence and feature analysis processing to generate a forgery detection report.

[0078] Specifically, the above steps include:

[0079] The tampering detection results, screenshot source detection results, price anomaly detection results, and marked high-risk screenshots are used as multi-dimensional detection results (including ELA difference map, noise inconsistency area, price mismatch point, etc.) and input into the preset large language model.

[0080] The system performs semantic analysis and confidence quantification on the multi-dimensional detection results using a pre-defined large language model, outputting semantic feature analysis results (e.g., "abnormal noise pattern in price area, suspected splicing") and confidence scores (e.g., authenticity score 0-100%). Based on the semantic feature analysis results and confidence scores, a forgery detection report is generated. Figure 4 The illustration shows a schematic diagram of the content of a forged detection report provided in an embodiment of the present invention. The report includes overall confidence level, detailed feature analysis (such as specific tampering points), and action suggestions to help staff make quick decisions.

[0081] Ultimately, the falsified test reports are provided to staff for analysis and decision-making, supporting both manual review and platform decisions (such as refusing to upload).

[0082] In addition, this embodiment of the invention also provides a device for detecting forgery of screenshots of hotel platform pages. The device includes: a processor and a memory; the memory is used to store one or more program instructions; the processor is used to run one or more program instructions to perform the steps of the forgery detection method for screenshots of hotel platform pages as described above.

[0083] In addition, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the forgery detection method for hotel platform page screenshots as described above.

[0084] In addition, this embodiment of the invention also provides a computer program product, which includes computer program instructions that, when executed by a processor, implement the steps of the forgery detection method for hotel platform page screenshots as described above. Figure 2 The diagram illustrates the architecture of a computer program product provided in an embodiment of the present invention. In this architecture, screenshots uploaded by users first enter the data input layer and are then distributed in parallel to three detection modules. Each module outputs detection results (such as traces of tampering), which are then integrated and input into a large model for comprehensive analysis, ultimately generating a report to support decision-making.

[0085] Compared with the prior art, the present invention has the following technical advantages:

[0086] This invention effectively improves the accuracy and real-time performance of screenshot forgery detection, achieves accurate identification of tampering traces, and reduces the response time to the second level.

[0087] This invention can effectively prevent the risk of fraudulent orders, detect malicious tampering as early as possible, and thus reduce economic losses to the platform and false price comparisons.

[0088] This invention provides quantitative decision support: confidence and feature reports generated by large models help staff make quick judgments and improve review efficiency.

[0089] This invention integrates physical and digital trace detection of images to form a closed-loop mechanism, promoting the healthy development of App functions and enhancing platform security.

[0090] This invention reduces operating costs: automation replaces manual review, making it suitable for scenarios with massive uploads.

[0091] In summary, this invention innovatively integrates EXIF, ELA, noise, and copy-move analysis to comprehensively identify physical and digital traces in hotel screenshots, achieving high-precision forgery detection. By combining OCR, template matching, non-maximum suppression, and content consistency checks, it forms a dedicated forgery detection method for hotel price comparison scenarios, overcoming the blind spots of traditional methods in detecting tampering in specific areas. Furthermore, this invention introduces a large language model, inputting multi-dimensional detection results into a large model for semantic analysis and confidence quantification, forming a closed-loop decision-making mechanism to enhance the intelligence and practicality of the detection.

[0092] In this embodiment of the invention, the processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in this embodiment of the invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in this embodiment of the invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The processor reads information from the storage medium and, in conjunction with its hardware, completes the steps of the above methods. The storage medium can be memory, for example, volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM).The storage media described in the embodiments of this invention are intended to include, but are not limited to, these and any other suitable types of memory. Those skilled in the art will recognize that the functions described in the above examples can be implemented using a combination of hardware and software. When applied software, the corresponding functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of computer programs from one place to another. Storage media can be any available medium accessible to general-purpose or special-purpose computers. Although the invention has been described in detail above with general description and specific embodiments, modifications or improvements can be made to it, which will be apparent to those skilled in the art. Therefore, such modifications or improvements made without departing from the spirit of the invention are all within the scope of protection claimed by the invention.

[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, or alterations made by those skilled in the art using the disclosed technical content shall fall within the protection scope of the present invention.

Claims

1. A method for detecting forgery of a hotel platform page screenshot, characterized in that, The method includes: Obtain screenshots of the hotel platform pages to be tested; Multi-dimensional forgery detection is performed on the screenshot to be detected to obtain multi-dimensional detection results. The multi-dimensional forgery detection includes image tampering detection, screenshot source detection, and price area anomaly detection. The multi-dimensional detection results are input into a preset large language model for confidence and feature analysis to generate a forgery detection report.

2. The method for detecting fraud against a hotel platform page screenshot according to claim 1, wherein, The screenshot to be detected is subjected to multi-dimensional forgery detection, resulting in multi-dimensional detection results, including: The screenshot to be detected is subjected to image tampering detection to obtain the tampering detection result; If the tampering detection result indicates that there are traces of tampering, then the screenshot to be detected will be marked as a high-risk screenshot; If the tampering detection result is no trace of tampering, then the screenshot to be detected is subjected to a screenshot source authenticity detection to obtain the screenshot source detection result; If the screenshot source detection result indicates that the source is not genuine, then the screenshot to be detected will be marked as a high-risk screenshot; If the screenshot source detection result indicates that the source is genuine, then price area anomaly detection is performed on the screenshot to be detected to obtain the price anomaly detection result; If the price anomaly detection result indicates the presence of a price anomaly, the screenshot to be detected will be marked as a high-risk screenshot. If the price anomaly detection result is that there is no price anomaly, then the screenshot to be detected will not be marked as a high-risk screenshot.

3. The method for detecting fraud against a hotel platform page screenshot according to claim 2, wherein, The screenshot to be detected is subjected to image tampering detection to obtain tampering detection results, including: Read the EXIF ​​metadata of the screenshot to be detected, and detect whether the EXIF ​​metadata contains image editing software tag information, wherein the image editing software tag information includes the image editing software name and / or image editing software version information; If the EXIF ​​metadata contains image editing software tag information, the tampering detection result indicates that there are traces of tampering. The screenshot to be detected is recompressed using JPEG to obtain a recompressed screenshot; Calculate the pixel differences between the recompressed screenshot and the original screenshot to be detected, and generate a visualized ELA difference map; Detect whether there are regions with inconsistent compression history in the ELA difference map; If there are areas of inconsistent compression history in the ELA difference map, the tampering detection result indicates that there are traces of tampering. A consistency analysis of the noise patterns in each region of the screenshot to be detected is performed to detect whether there are regions with abnormal noise patterns in the screenshot to be detected. If there is an abnormal noise pattern area in the screenshot to be detected, the tampering detection result is considered to have been tampered with; Feature matching is performed within the screenshot to be detected to detect whether there are two or more regions in the screenshot to be detected with a similarity greater than a preset similarity threshold. If there are two or more regions in the screenshot to be detected that have a similarity greater than a preset similarity threshold, the tampering detection result is that there are traces of tampering. If the EXIF ​​metadata does not contain image editing software tag information, and the ELA difference map does not contain areas with inconsistent compression history, and the screenshot to be detected does not contain areas with abnormal noise patterns, and the screenshot to be detected does not contain more than two areas with similarity greater than a preset similarity threshold, then the tampering detection result is no tampering trace.

4. The method for detecting fraud against a hotel platform page screenshot according to claim 2, wherein, The screenshot to be detected is subjected to a source authenticity detection test to obtain the screenshot source detection result, including: Perform a Fourier transform on the screenshot to be detected to obtain the frequency domain features of the screenshot; Based on the frequency domain characteristics of the screenshot, determine whether there are traces of digital generation in the screenshot to be detected; If the screenshot to be detected contains traces of digital generation, then the screenshot source detection result is that the source is not genuine; If there are no traces of digital generation in the screenshot to be detected, then the screenshot source detection result is that the source is genuine.

5. The method for detecting fraud against a hotel platform page screenshot according to claim 2, wherein, The screenshot to be detected is subjected to price region anomaly detection, and the price anomaly detection results are obtained, including: By using ORC to identify and preset hotel price data templates, candidate price areas are located in the screenshot to be detected. Non-maximum suppression is applied to the candidate price region to remove duplicate boxes, resulting in the final price region. Image tampering detection is performed on the final price area to obtain the price area tampering detection result; If there are more than two final price regions, then a content consistency check is performed on each final price region to obtain the price consistency check result.

6. The method for detecting fraud against a hotel platform page screenshot according to claim 2, wherein, The multi-dimensional detection results are input into a preset large language model for confidence and feature analysis to generate a forgery detection report, including: The tampering detection result, the screenshot source detection result, the price anomaly detection result, and the high-risk screenshot are input as multi-dimensional detection results into a preset large language model; The multi-dimensional detection results are subjected to semantic analysis and confidence quantification by the preset large language model to obtain semantic feature analysis results and confidence scores; Based on the semantic feature analysis results and the confidence score, a forgery detection report is generated.

7. A device for fake detection against hotel platform page screenshots, characterized by, The device includes: a processor and a memory; The memory is used to store one or more program instructions; The processor is configured to run one or more program instructions to perform the steps of the forgery detection method for hotel platform page screenshots as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the forgery detection method for hotel platform page screenshots as described in any one of claims 1 to 6.

9. A computer program product, characterized in that, The computer program product includes computer program instructions that, when executed by a processor, implement the steps of the forgery detection method for hotel platform page screenshots as described in any one of claims 1 to 6.

Images