Video content detection method, apparatus, device, and storage medium

Abstract

The application discloses a video content detection method, device and equipment and a storage medium, and the method comprises the following steps: determining a video frame set corresponding to a target video frame in a video frame sequence; determining a difference image between the video frames in the video frame set; determining a to-be-detected image of the target video frame according to the difference image; and detecting the to-be-detected image to obtain a detection result of the target video frame. The application improves the accuracy and comprehensiveness of video detection.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a video content detection method, apparatus, device, and storage medium. Background Technology

[0002] With the rapid development of the internet across various fields, a vast amount of content and information is disseminated online, among which video content is the most common form of information transmission. To build a healthy online ecosystem and ensure the security of video content, the review and approval of video content is crucial.

[0003] Existing video content detection methods identify static images within the video content to determine whether it meets playback standards. However, video content consists of multiple consecutive frames. Due to the persistence of vision, the combination of images over a period of time or in successive frames may generate new meanings, leading to errors in the video content detection results. Summary of the Invention

[0004] The main objective of this invention is to provide a video content detection method, apparatus, device, and storage medium, aiming to solve the problem of how to improve the accuracy of video content detection.

[0005] To achieve the above objectives, the present invention provides a video content detection method, which includes the following steps:

[0006] Determine the set of video frames corresponding to the target video frame in the video frame sequence;

[0007] Determine the difference images between video frames in the video frame set;

[0008] Based on the difference image, determine the image to be detected corresponding to the target video frame;

[0009] The image to be detected is then detected to obtain the detection result of the target video frame.

[0010] Optionally, the step of determining the set of video frames corresponding to the target video frame in the video frame sequence includes:

[0011] Determine the video frame sequence of the video to be detected;

[0012] Based on the frame order of the video frame sequence, the video frames in the video frame sequence are sequentially determined as target video frames, and the set of video frames corresponding to the target video frames is determined, until the target video frame is the last video frame of the video frame sequence.

[0013] Optionally, the step of determining the set of video frames corresponding to the target video frame includes:

[0014] The number of frames in the video frame set is determined based on the video frame rate and time domain of the video to be detected.

[0015] Based on the number of frames, consecutive video frames containing the target video frame are selected as the video frame set corresponding to the target video frame.

[0016] Optionally, the step of determining the image to be detected corresponding to the target video frame based on the difference image includes:

[0017] Determine the difference data of the difference images;

[0018] Determine the background data of the target video frame;

[0019] The image to be detected is determined based on all the difference data and the background data.

[0020] Optionally, the step of determining the background data of the target video frame includes:

[0021] Determine the target data of a preset type in the target video frame;

[0022] The background data is determined based on the target data.

[0023] Optionally, the step of determining the difference images between video frames in the video frame set includes:

[0024] Determine the similarity between any two video frames in the set of video frames;

[0025] If the similarity is less than a preset threshold, then the image information and / or text information of the two video frames whose similarity is less than the preset threshold are determined;

[0026] The difference image is obtained by comparing the image information and / or text information of two video frames.

[0027] Optionally, the step of detecting the image to be detected and obtaining the detection result of the target video frame includes:

[0028] The image to be detected is input into the text recognition model and the image classification model respectively to obtain text information and image information;

[0029] The detection result of the target video frame is determined based on the text information and the image information.

[0030] To achieve the above objectives, the present invention also provides a video content detection device, the device comprising:

[0031] The acquisition module is used to determine the set of video frames corresponding to the target video frame in the video frame sequence;

[0032] The determining module is used to determine the difference images between video frames in the video frame set;

[0033] The reconstruction module is used to determine the image to be detected corresponding to the target video frame based on the difference image;

[0034] The detection module is used to detect the image to be detected and obtain the detection result of the target video frame.

[0035] To achieve the above objectives, the present invention also provides a video content detection device, the video content detection device including a memory, a processor, and a video content detection program stored in the memory and executable on the processor, wherein the video content detection program, when executed by the processor, implements the various steps of the video content detection method as described above.

[0036] To achieve the above objectives, the present invention also provides a computer-readable storage medium storing a video content detection program, which, when executed by a processor, implements the various steps of the video content detection method described above.

[0037] This invention provides a video content detection method, apparatus, device, and storage medium. The method involves: determining the set of video frames corresponding to a target video frame in a video frame sequence; determining the difference images between video frames in the set; determining the image to be detected corresponding to the target video frame based on the difference images; and detecting the image to be detected to obtain the detection result of the target video frame. By determining the image to be detected corresponding to the target video frame, and since the image to be detected includes difference data and background data, the detection of the image to be detected not only detects the content of a single frame of the target video frame but also detects the difference images of adjacent video frames. This avoids the generation of new meanings after multiple consecutive video frames are combined, thus improving the accuracy and comprehensiveness of video detection. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the hardware structure of the video content detection device according to an embodiment of the present invention;

[0039] Figure 2 This is a flowchart illustrating the first embodiment of the video content detection method of the present invention;

[0040] Figure 3 This is a schematic diagram illustrating the relationship between the difference image and the image to be detected in the video content detection method of the present invention;

[0041] Figure 4 This is a detailed flowchart of step S10 in the second embodiment of the video content detection method of the present invention;

[0042] Figure 5 This is a schematic diagram of the video frame sequence and time offset of the video content detection method of the present invention;

[0043] Figure 6 This is a schematic diagram illustrating the relationship between the video frame set and the image to be detected in the video content detection method of the present invention;

[0044] Figure 7 This is a detailed flowchart of step S20 in the third embodiment of the video content detection method of the present invention.

[0045] Figure 8 This is a schematic diagram of the logical structure of the video content detection device according to an embodiment of the present invention.

[0046] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0047] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0048] The main solution of this invention is: to determine the set of video frames corresponding to the target video frame in the video frame sequence; to determine the difference images between the video frames in the set of video frames; to determine the image to be detected of the corresponding target video frame based on the difference images; and to detect the image to be detected to obtain the detection result of the target video frame.

[0049] By identifying the image to be detected corresponding to the target video frame, and since the image to be detected includes difference data and background data, the detection of the image to be detected not only detects the content of a single frame of the target video frame, but also detects the difference images of adjacent video frames of the target video frame. This avoids the generation of new meanings after multiple consecutive video frames are combined, thus improving the accuracy and comprehensiveness of video detection.

[0050] As one implementation solution, video content inspection equipment can, for example... Figure 1 As shown.

[0051] The present invention relates to a video content detection device, which includes: a processor 101, such as a CPU, a memory 102, and a communication bus 103. The communication bus 103 is used to enable communication between these components.

[0052] Memory 102 can be high-speed RAM or stable memory (non-volatile memory), such as disk storage. Figure 1As shown, the memory 102, which is a computer-readable storage medium, may include a video content detection program; and the processor 101 may be used to invoke the video content detection program stored in the memory 102 and perform the following operations:

[0053] Determine the set of video frames corresponding to the target video frame in the video frame sequence;

[0054] Determine the difference images between video frames in the video frame set;

[0055] Based on the difference image, determine the image to be detected corresponding to the target video frame;

[0056] The image to be detected is then detected to obtain the detection result of the target video frame.

[0057] Optionally, the processor 101 can be used to invoke the video content detection program stored in the memory 102 and perform the following operations:

[0058] Determine the video frame sequence of the video to be detected;

[0059] Based on the frame order of the video frame sequence, the video frames in the video frame sequence are sequentially determined as target video frames, and the set of video frames corresponding to the target video frames is determined, until the target video frame is the last video frame of the video frame sequence.

[0060] Optionally, the processor 101 can be used to invoke the video content detection program stored in the memory 102 and perform the following operations:

[0061] The number of frames in the video frame set is determined based on the video frame rate and time domain of the video to be detected.

[0062] Based on the number of frames, consecutive video frames containing the target video frame are selected as the video frame set corresponding to the target video frame.

[0063] Optionally, the processor 101 can be used to invoke the video content detection program stored in the memory 102 and perform the following operations:

[0064] Determine the difference data of the difference images;

[0065] Determine the background data of the target video frame;

[0066] The image to be detected is determined based on all the difference data and the background data.

[0067] Optionally, the processor 101 can be used to invoke the video content detection program stored in the memory 102 and perform the following operations:

[0068] Determine the target data of a preset type in the target video frame;

[0069] The background data is determined based on the target data.

[0070] Optionally, the processor 101 can be used to invoke the video content detection program stored in the memory 102 and perform the following operations:

[0071] Determine the similarity between any two video frames in the set of video frames;

[0072] If the similarity is less than a preset threshold, then the image information and / or text information of the two video frames whose similarity is less than the preset threshold are determined;

[0073] The difference image is obtained by comparing the image information and / or text information of two video frames.

[0074] Optionally, the processor 101 can be used to invoke the video content detection program stored in the memory 102 and perform the following operations:

[0075] The image to be detected is input into the text recognition model and the image classification model respectively to obtain text information and image information;

[0076] The detection result of the target video frame is determined based on the text information and the image information.

[0077] Based on the hardware architecture of the video content detection device described above, an embodiment of the video content detection method of the present invention is proposed.

[0078] Reference Figure 2 , Figure 2 This is a first embodiment of the video content detection method of the present invention, which includes the following steps:

[0079] Step S10: Determine the set of video frames corresponding to the target video frame in the video frame sequence;

[0080] Optionally, a video frame sequence corresponding to the video to be detected is determined, wherein the video frame sequence is a series of consecutive video frames arranged in chronological order. Optionally, the video frame sequence includes video frame S1, video frame S2, ..., video frame Sn, wherein video frame S1 is the first video frame and video frame Sn is the last video frame. The video frames in the video frame sequence are sequentially used as target video frames, and the target video frame is the video frame currently being detected.

[0081] Determine the set of video frames corresponding to the target video frame. Optionally, the set of video frames can be a set of consecutive video frames that are adjacent to the target video frame for a preset number of frames. For example, the preset number of frames in the set of video frames is 3, the target video frame is the i-th frame, and the set of video frames corresponding to the target video frame includes the (i-1)-th video frame, the i-th target video frame, and the (i+1)-th video frame.

[0082] Step S20: Determine the difference images between video frames in the video frame set.

[0083] Optionally, determine the difference image between every two video frames in the video frame set, optionally, such as Figure 3 As shown, when the video frame set of the target video frame S2 includes video frame S1, target video frame S2, and video frame S3, a difference image Diff-s12 is determined between video frame S1 and target video frame S2, a difference image Diff-s23 is determined between target video frame S2 and video frame S3, and a difference image Diff-s13 is determined between video frame S1 and video frame S3. Optionally, the difference image includes difference data between every two video frames, and optionally, the difference data includes text information and / or image information and / or contour information, etc.

[0084] Optionally, the similarity between any two video frames in the video frame set is determined. If the similarity is less than a preset threshold, it indicates a significant difference between the two video frames. In this case, the image information and / or text information of the two video frames with similarity less than the preset threshold are determined. The image information and / or text information of the two video frames are compared to obtain a difference image. Optionally, the image information includes the contour information of the moving object. Optionally, the similarity between the two video frames is calculated by analyzing features such as pixel changes, brightness, contrast, and structure. The similarity can be calculated using SSIM (Structural Similarity) and / or mean squared error between any two video frames in the video frame set.

[0085] Step S30: Based on the difference image, determine the image to be detected for the corresponding target video frame.

[0086] Optionally, the image to be detected includes difference data from all difference images in the video frame set. Alternatively, the image to be detected includes difference data from all difference images in the video frame set, as well as background data from the target video frame.

[0087] Optionally, the difference data of the difference images are determined, and the image to be detected is determined based on all the difference data and the background data Bak-2 of the target video frame S2. Optionally, target data of a preset type is determined in the target video frame; based on the target data, background data is determined, wherein the preset type is the type of sensitive information of interest for image detection, such as a square pattern or square text. Optionally, the background data is image information other than the difference data between the difference images. The difference data of difference images Diff-s12, Diff-s23, and Diff-s13 are determined respectively, and the image to be detected corresponding to the target video frame is determined based on all the difference data and the background data Bak-2.

[0088] Step S40: Detect the image to be detected to obtain the detection result of the target video frame.

[0089] Optionally, the image to be detected is input into both a text recognition model and an image classification model for detection to obtain text information and / or image information in the image; based on the text information and / or image information, the detection result of the target video frame is determined. Optionally, the text recognition model can be an OCR (Optical Character Recognition) model. Optionally, the image classification model can be a multi-scene classification and detection model built on deep networks such as YOLO, SSD, MTCNN, and ArcFace.

[0090] Optionally, if the text information is preset text information or the image information is preset image information, the detection result of the target video frame is determined to be unsuccessful; if the text information is not preset text information and the image information is not preset image information, the detection result of the target video frame is determined to be successful.

[0091] Optionally, if the video to be tested contains video frames that fail the review, the video is determined to have failed the review, and cannot be published to the platform or played. If the video to be tested does not contain any video frames that fail the review, the video is determined to have passed the review, and can be published to the platform and played on the platform.

[0092] In this embodiment, the technical solution involves determining the set of video frames corresponding to the target video frame in a video frame sequence; determining the difference images between video frames in the set; determining the image to be detected for the corresponding target video frame based on the difference images; and detecting the image to be detected to obtain the detection result of the target video frame. By determining the image to be detected corresponding to the target video frame, and since the image to be detected includes difference data and background data, the detection of the image to be detected not only detects the content of a single frame of the target video frame but also detects the difference images of adjacent video frames. This avoids the generation of new meanings after multiple consecutive video frames are combined, thus improving the accuracy and comprehensiveness of video detection.

[0093] Reference Figure 4 , Figure 4 This is a second embodiment of the video content detection method of the present invention. Based on the first embodiment, step S10 includes:

[0094] Step S11: Determine the video frame sequence of the video to be detected;

[0095] Step S12: Based on the frame order of the video frame sequence, determine the video frames in the video frame sequence as target video frames in turn, and determine the set of video frames corresponding to the target video frames, until the target video frame is the last video frame of the video frame sequence.

[0096] Optionally, the video frame sequence of the video to be detected is video frame S1, video frame S2, ..., video frame Sn, where video frame S1 is the first video frame and video frame Sn is the last video frame. Based on the frame order of the video frame sequence, the video frames in the video frame sequence are sequentially taken as target video frames, and the target video frames are the video frames currently being detected. Optionally, the set of video frames corresponding to the target video frames is a set of multiple consecutive video frames.

[0097] Optionally, the number of frames in the video frame set is determined based on the video frame rate and time domain of the video to be detected; and consecutive video frames containing the target video frame are selected as the video frame set corresponding to the target video frame based on the number of frames.

[0098] Optionally, based on the video frame rate of the video to be detected and the time offset between adjacent video frames, the video frames in the video frame sequence are time-stamped to determine the frame rate of the video to be detected as F. Taking the start position of the video or a specified moment as the starting point, the relative time offset of each frame is marked as 1 / F seconds, and the time-stamping is performed sequentially on video frames S1, S2, ... and Sn. Figure 5 As shown, the time point corresponding to video frame S1 is T1, the time point corresponding to video frame S2 is T2, ..., the time point corresponding to video frame Sn is Tn, and the relative time offset between each time point is 1 / F seconds.

[0099] Optionally, the time domain is the time range corresponding to the set of video frames. Time domain = number of frames * (1 / video frame rate). For example, if the video frame rate is 24 / s, the offset time between the first and last video frames before and after the target video frame is selected, and the sum of the offset times is determined as the time domain, i.e., 1 / 24s + 1 / 24s + 1 / 24s = 3 / 24s, where 3 / 24s is one time domain.

[0100] The number of frames in the video frame set is determined based on the video frame rate and the time domain. Optionally, the number of frames = time domain * video frame rate. For example, if the video frame rate is 24 / s and the time domain is 3 / 24s, then the number of frames is 3 / 24s * 24 / s = 3, where the time domain 3 / 24s represents the dwell time of the target video frame and the dwell time of the two preceding and following video frames when the video frame rate is 24 / s.

[0101] The set of consecutive video frames containing the target video frame is selected as the video frame set corresponding to the target video frame based on the frame number. Optionally, when the target video frame is not the first or last video frame, the video frame set is determined based on the frame number 2n+1, that is, the video frame set contains 2n+1 frames. The video frame set includes the target video frame and n adjacent video frames before and after the target video frame. For example, when the video frame set contains 3 frames, the video frame set includes the target video frame and the video frames of the frame before and after the target video frame. Figure 6 As shown, the set of video frames corresponding to video frame S2 is video frame S1, video frame S2 and video frame S3, where the image to be detected corresponding to video frame Si is Si-new, 1 < i < n.

[0102] Optionally, when the target video frame is the first video frame, the video frame set is determined based on the frame number 2n+1. Since there are no video frames before the first video frame, the video frame set includes the target video frame and n consecutive video frames following the target video frame. For example, when the number of video frames in the video frame set is 3, the video frame set includes the target video frame and the video frames of the frame following the target video frame, such as... Figure 5 As shown, the set of video frames corresponding to video frame S1 is video frame S1 and video frame S2, where the image to be detected corresponding to video frame S1 is S1-new.

[0103] Optionally, when the target video frame is the last video frame, the video frame set is determined based on the frame number 2n+1. Since there are no video frames after the last video frame, the video frame set includes the target video frame and n consecutive video frames preceding the target video frame. For example, when the number of video frames in the video frame set is 3, the video frame set includes the target video frame and the video frame of the frame preceding the target video frame, such as... Figure 5As shown, the set of video frames for video frame Sn is video frame Sn-1 and video frame Sn, where the image to be detected corresponding to video frame Sn is Sn-new.

[0104] In this embodiment, the technical solution involves determining a video frame sequence of the video to be detected; based on the frame order of the video frame sequence, sequentially identifying the video frames in the video frame sequence as target video frames; and determining the set of video frames corresponding to the target video frames. By determining the set of video frames corresponding to the target video frames in the video frame sequence, it is easier to subsequently determine the difference data of the difference images, thereby improving the comprehensiveness of the detection of the target video frames.

[0105] Reference Figure 7 , Figure 7 In a third embodiment of the video content detection method of the present invention, based on the first or second embodiment, step S30 includes:

[0106] Step S31: Determine the difference data of the difference image;

[0107] Step S32: Determine the background data of the target video frame;

[0108] Step S33: Determine the image to be detected based on all the difference data and the background data.

[0109] like Figure 3 As shown, when the video frame set of the target video frame S2 includes video frame S1, target video frame S2, and video frame S3, the difference image Diff-s12 between video frame S1 and target video frame S2 is determined, the difference image Diff-s23 between target video frame S2 and video frame S3 is determined, and the difference image Diff-s13 between video frame S1 and video frame S3 is determined. Optionally, the difference data of difference images Diff-s12, Diff-s23, and Diff-s13 are determined respectively, and the image to be detected corresponding to the target video frame is determined based on all the difference data and background data Bak-2.

[0110] Optionally, target data of a preset type is determined in the target video frame; based on the target data, background data is determined, wherein the preset type is the type of sensitive information of interest for image detection, such as a square pattern or square text. Optionally, the background data is image information other than the difference data between the different images.

[0111] Optionally, all the difference data is reconstructed with the background data of the target video frame to obtain a difference image. The background data of the target video frame is the image portion outside the difference data in the time domain. For example, circular patterns and "square" text are sensitive information. Detecting the target video frame separately only detects the circular image within it. However, the image to be detected includes not only the difference data of difference images Diff-s12, Diff-s23, and Diff-s13 (e.g., the difference data corresponding to the difference images are square patterns and "square" text), but also the background data of target video frame S2, i.e., the circular pattern. Finally, these are reconstructed, and the final difference image includes the square pattern, "square" text, and circular pattern. Figure 3 As shown, the difference data and background data are recombined to improve the accuracy and comprehensiveness of target video frame detection.

[0112] In this embodiment, the difference data of the difference image is determined; the background data of the target video frame is determined; and the image to be detected is determined based on all the difference data and background data. By determining the background data and difference data, and recombining the difference data and background data, the image to be detected is determined. The image to be detected includes the background data of the target video frame and the difference data of the difference image. This avoids the generation of new meanings after combining multiple consecutive video frames, and improves the accuracy and comprehensiveness of target video frame detection.

[0113] Reference Figure 8 The present invention also provides a video content detection device, the device comprising:

[0114] The acquisition module 100 is used to determine the set of video frames corresponding to the target video frame in the video frame sequence;

[0115] The determining module 200 is used to determine the difference images between video frames in the video frame set;

[0116] The reconstruction module 300 is used to determine the image to be detected corresponding to the target video frame based on the difference image;

[0117] The detection module 400 is used to detect the image to be detected and obtain the detection result of the target video frame.

[0118] Optionally, the step of determining the set of video frames corresponding to the target video frame in the video frame sequence includes:

[0119] Determine the video frame sequence of the video to be detected;

[0120] Based on the frame order of the video frame sequence, the video frames in the video frame sequence are sequentially determined as target video frames, and the set of video frames corresponding to the target video frames is determined, until the target video frame is the last video frame of the video frame sequence.

[0121] Optionally, the step of determining the set of video frames corresponding to the target video frame includes:

[0122] The number of frames in the video frame set is determined based on the video frame rate and time domain of the video to be detected.

[0123] Based on the number of frames, consecutive video frames containing the target video frame are selected as the video frame set corresponding to the target video frame.

[0124] Optionally, the step of determining the image to be detected corresponding to the target video frame based on the difference image includes:

[0125] Determine the difference data of the difference images;

[0126] Determine the background data of the target video frame;

[0127] The image to be detected is determined based on all the difference data and the background data.

[0128] Optionally, the step of determining the background data of the target video frame includes:

[0129] Determine the target data of a preset type in the target video frame;

[0130] The background data is determined based on the target data.

[0131] Optionally, the step of determining the difference images between video frames in the video frame set includes:

[0132] Determine the similarity between any two video frames in the set of video frames;

[0133] If the similarity is less than a preset threshold, then the image information and / or text information of the two video frames whose similarity is less than the preset threshold are determined;

[0134] The difference image is obtained by comparing the image information and / or text information of two video frames.

[0135] Optionally, the step of detecting the image to be detected and obtaining the detection result of the target video frame includes:

[0136] The image to be detected is input into the text recognition model and the image classification model respectively to obtain text information and image information;

[0137] The detection result of the target video frame is determined based on the text information and the image information.

[0138] The present invention also provides a video content detection device, the video content detection device including a memory, a processor, and a video content detection program stored in the memory and executable on the processor, wherein when the video content detection program is executed by the processor, it implements the various steps of the video content detection method described in the above embodiments.

[0139] The present invention also provides a computer-readable storage medium storing a video content detection program, which, when executed by a processor, implements the various steps of the video content detection method described in the above embodiments.

[0140] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0141] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, system, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, system, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, system, article, or apparatus that includes that element.

[0142] Through the above description of the embodiments, those skilled in the art can clearly understand that the systems described in the embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, parking management device, air conditioner, or network device, etc.) to execute the systems described in the various embodiments of the present invention.

[0143] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A video content detection method, characterized in that, The method includes: Determine the video frame sequence of the video to be detected; Based on the frame order of the video frame sequence, video frames in the video frame sequence are sequentially determined as target video frames, and a set of video frames corresponding to the target video frames is determined, until the target video frame is the last video frame of the video frame sequence; wherein, when the target video frame is not the first or last video frame, the set of video frames includes the target video frame and n video frames adjacent to the target video frame; when the target video frame is the first video frame, the set of video frames includes the target video frame and n consecutive video frames adjacent to the target video frame; when the target video frame is the last video frame, the set of video frames includes the target video frame and n consecutive video frames adjacent to the target video frame. Determine the difference image between every two video frames in the video frame set corresponding to the target video frame, wherein the difference image includes the difference data between every two video frames; Determine the difference data of the difference images; Determine the background data of the target video frame; Based on all the difference data and the background data, determine the image to be detected corresponding to the target video frame; The image to be detected is detected to obtain the detection result of the target video frame, and the detection result includes whether the review is passed or not.

2. The video content detection method as described in claim 1, characterized in that, The step of determining the set of video frames corresponding to the target video frame further includes: The number of frames in the video frame set is determined based on the video frame rate of the video to be detected and the time range corresponding to the video frame set. Based on the number of frames, consecutive video frames containing the target video frame are selected as the video frame set corresponding to the target video frame.

3. The video content detection method as described in claim 1, characterized in that, The step of determining the background data of the target video frame includes: Determine the target data of a preset type in the target video frame; The background data is determined based on the target data.

4. The video content detection method as described in claim 1, characterized in that, The step of determining the difference image between every two video frames in the set of video frames corresponding to the target video frame includes: Determine the similarity between any two video frames in the set of video frames; If the similarity is less than a preset threshold, then the image information and / or text information of the two video frames whose similarity is less than the preset threshold are determined; The difference image is obtained by comparing the image information and / or text information of two video frames.

5. The video content detection method as described in claim 1, characterized in that, The step of detecting the image to be detected and obtaining the detection result of the target video frame includes: The image to be detected is input into the text recognition model and the image classification model respectively to obtain text information and image information; The detection result of the target video frame is determined based on the text information and the image information.

6. A video content detection device, characterized in that, The device includes: An acquisition module is used to determine the video frame sequence of the video to be detected; based on the frame order of the video frame sequence, it sequentially determines the video frames in the video frame sequence as target video frames, and determines the video frame set corresponding to the target video frame, until the target video frame is the last video frame of the video frame sequence; wherein, when the target video frame is not the first or last video frame, the video frame set includes the target video frame and n adjacent video frames before and after the target video frame; when the target video frame is the first video frame, the video frame set includes the target video frame and n consecutive video frames after the target video frame; when the target video frame is the last video frame, the video frame set includes the target video frame and n consecutive video frames before the target video frame. The determining module is used to determine the difference image between every two video frames in the video frame set corresponding to the target video frame, wherein the difference image includes the difference data between every two video frames. A reconstruction module is used to determine the difference data of the difference image; determine the background data of the target video frame; and determine the image to be detected corresponding to the target video frame based on all the difference data and the background data. The detection module is used to detect the image to be detected and obtain the detection result of the target video frame, wherein the detection result includes whether the review is passed or not.

7. A video content detection device, characterized in that, The video content detection device includes a memory, a processor, and a video content detection program stored in the memory and executable on the processor. When the video content detection program is executed by the processor, it implements the various steps of the video content detection method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a video content detection program, which, when executed by a processor, implements the steps of the video content detection method as described in any one of claims 1-5.

Images