E-commerce live broadcast management system

Abstract

The invention relates to an e-commerce live broadcast management system, and the system comprises a signal extraction mechanism which is used for carrying out the image signal extraction operation of a live broadcast room in which a live broadcast user reports that human body privacy parts are leaked, so as to randomly extract a single-frame live broadcast image from multi-frame live broadcast images corresponding to a plurality of continuous moments, and to output the single-frame live broadcast image as a monitoring image; and the code printing processing component is used for performing code printing processing on a certain reference image block existing in the customized filtering image when receiving an object identification instruction. According to the invention, image signal extraction operation can be carried out on the live broadcast room in which the live broadcast user reports that the human body privacy part is leaked, and the randomly extracted single-frame live broadcast picture is used as the monitoring picture to carry out targeted intelligent analysis on the existence of the human body privacy part; and when the analysis result judges that the human body privacy part exists, directional coding processing is performed on the live broadcast picture, so that the real-time performance of shielding processing of the human body privacy part of the live broadcast picture is improved.

Description

Technical Field

[0001] The present invention relates to the field of live broadcast strategy processing, and in particular to an e-commerce live broadcast management system. Background Art

[0002] With the rapid development of Internet technology, live streaming has become one of the most popular forms of entertainment. However, behind this seemingly prosperous market, there are many problems. These behaviors not only damage the privacy of users, but also bring negative impacts to the entire live streaming industry.

[0003] Therefore, it is necessary to quickly code some live broadcast images that expose private parts of the human body to avoid leakage of live broadcast images that expose private parts of the human body. However, in the prior art, whether the live broadcast images expose private parts of the human body is generally reviewed on the live broadcast server side. This review processing mode is time-consuming and labor-intensive on the one hand, and on the other hand, the real-time broadcast image content is often not responded to in a timely manner, resulting in the possibility of live broadcast images that expose private parts of the human body still leaking.

[0004] The prior art disclosed includes: the invention "electronic media advertising content intelligent identification system" with application publication number CN110119788A, which includes: a perception unit, which is used to obtain the advertising screen image played on the electronic media screen; a deep learning unit, which is used to receive the advertising screen image obtained by the perception unit and identify the advertising screen image obtained by the perception unit; an advertising screen image database, which is used to store a specific type of advertising screen image identified by the deep learning unit; an action unit, which obtains the recognition result given by the deep learning unit and controls the power supply of the electronic media screen according to preset rules. The invention can intelligently identify images involving pornography, explosions and terrorism, and when an illegal image is identified, the electronic media advertising machine can be automatically powered off and shut down in time to prevent the possibility of causing negative social impact. The invention "Live Video Identification Method, Device, Equipment and Medium" with application publication number CN112801062A can first perform preliminary skin color detection based on brightness, effectively avoiding misjudgment caused by brightness, and further adjust and optimize the elliptical skin color model in combination with brightness, avoiding the effect of skin color detection affected by brightness, effectively reducing the false detection rate, and then automatically detecting illegal live broadcasts in live videos based on the improved elliptical skin color model. At the same time, since it is frame-by-frame detection, it also avoids misjudgment of illegal scenes that appear for a short time in traditional solutions due to the use of video frame training. Summary of the invention

[0005] In order to solve the technical problems in the prior art, according to the present invention, an e-commerce live broadcast management system is provided, the system comprising: The signal extraction mechanism is used to extract image signals from live broadcast rooms reported by live broadcast users to reveal private parts of human body, so as to randomly extract a single frame of live broadcast pictures from multiple frames of live broadcast pictures corresponding to multiple consecutive moments as monitoring picture output; An affine transformation device, connected to the signal extraction mechanism, for performing affine transformation processing on the received monitoring picture to obtain and output a corresponding affine transformation image; an edge sharpening device, connected to the affine transformation device, for performing edge sharpening processing on the received affine transformation image to obtain and output a corresponding edge sharpened image; a customized filtering device, connected to the edge sharpening device, for performing a salt and pepper noise removal action on the received edge sharpened image to obtain and output a customized filtered image, wherein the salt and pepper noise removal action is based on a median filtering algorithm; A targeted identification component is used to obtain the brightness values ​​corresponding to the pixels in the received customized filtered image, take each pixel in the customized filtered image as a central pixel, determine the brightness gradient of the central pixel based on the brightness values ​​of the central pixel and the pixels around it, and when the brightness gradient of the central pixel exceeds a limit, judge the central pixel as an edge pixel to obtain the edge pixels in the customized filtered image, and use the image blocks surrounded by the edge pixels in the customized filtered image as reference image blocks. When the contour of a reference image block in the customized filtered image matches the reference contour of the private part of the human body, an object identification instruction is issued; The coding processing component is connected to the targeted identification component and is used for coding the reference image block existing in the customized filtering image when receiving the object identification instruction.

[0006] An e-commerce live broadcast management system provided by the present invention can, on the basis of customized visual data processing, perform image signal extraction operations on the live broadcast rooms where live broadcast users report that private parts of the human body have been leaked, so as to randomly extract a single frame of live broadcast pictures from multiple frames of live broadcast pictures corresponding to multiple consecutive moments as a monitoring picture to perform targeted intelligent analysis of the existence of private parts of the human body, and perform targeted coding processing on the live broadcast pictures when the analysis results determine that private parts of the human body exist, thereby improving the real-time performance of the shielding processing of private parts of the human body in the live broadcast pictures. DETAILED DESCRIPTION

[0007] The following is a detailed description of an embodiment of the e-commerce live broadcast management system of the present invention.

[0008] Example 1

[0009] According to the e-commerce live broadcast management system shown in Embodiment 1 of the present invention, the system includes: The signal extraction mechanism is used to extract image signals from live broadcast rooms reported by live broadcast users to reveal private parts of human body, so as to randomly extract a single frame of live broadcast pictures from multiple frames of live broadcast pictures corresponding to multiple consecutive moments as monitoring picture output; For example, a signal extraction mechanism is used to extract image signals from a live broadcast room reported by a live broadcast user as having leaked private parts of the human body, so as to randomly extract a single frame of live broadcast pictures from multiple frames of live broadcast pictures corresponding to multiple consecutive moments as a monitoring picture output, including: the signal extraction mechanism includes a time service unit, an extraction execution unit and a picture output unit; An affine transformation device, connected to the signal extraction mechanism, for performing affine transformation processing on the received monitoring picture to obtain and output a corresponding affine transformation image; an edge sharpening device, connected to the affine transformation device, for performing edge sharpening processing on the received affine transformation image to obtain and output a corresponding edge sharpened image; a customized filtering device, connected to the edge sharpening device, for performing a salt and pepper noise removal action on the received edge sharpened image to obtain and output a customized filtered image, wherein the salt and pepper noise removal action is based on a median filtering algorithm; A targeted identification component is used to obtain the brightness values ​​corresponding to the pixels in the received customized filtered image, take each pixel in the customized filtered image as a central pixel, determine the brightness gradient of the central pixel based on the brightness values ​​of the central pixel and the pixels around it, and when the brightness gradient of the central pixel exceeds a limit, judge the central pixel as an edge pixel to obtain the edge pixels in the customized filtered image, and use the image blocks surrounded by the edge pixels in the customized filtered image as reference image blocks. When the contour of a reference image block in the customized filtered image matches the reference contour of the private part of the human body, an object identification instruction is issued; A coding processing component, connected to the targeted identification component, for performing coding processing on the reference image block existing in the customized filtered image upon receiving the object identification instruction; Wherein, the targeted identification component is further used to issue an object unidentified instruction when the contours of each reference image block in the customized filter image do not match the reference contours of the private parts of the human body; The coding processing component is further configured to not perform coding processing on the reference image block existing in the customized filtered image when receiving the object unrecognized instruction.

[0010] Example 2

[0011] An e-commerce live broadcast management system is shown according to Example 2 of the present invention.

[0012] Different from Example 1, the e-commerce live broadcast management system in Example 2 may further include: An optimization processing mechanism is connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, respectively, and is used to perform signal optimization processing on the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, so as to obtain respective optimized processed images corresponding to the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component; The signal optimization processing is performed on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, including: the first step of the optimization processing is to perform edge sharpening operation on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively; The signal optimization processing is performed on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, and the second step of optimization processing is to perform content enhancement operation on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively; Among them, respectively performing signal optimization processing on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, respectively, further includes: the third step of optimization processing respectively performs filtering processing operations on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component; Among them, signal optimization processing is performed on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, and also includes: the fourth step of optimization processing performs interpolation processing operations on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively.

[0013] Example 3

[0014] An e-commerce live broadcast management system is shown according to Example 3 of the present invention.

[0015] Different from Example 1, the e-commerce live broadcast management system in Example 3 may further include: A data display mechanism is connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, and is used to perform real-time display of each optimized processed image corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively; Wherein, the data display mechanism is connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, and is used to perform the instant display of each optimized processed image corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, including: the data display mechanism is a liquid crystal display device; And wherein, the data display mechanism is respectively connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, and is used to perform real-time display of each optimized processed image corresponding to the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, including: the data display mechanism is a split-screen display device.

[0016] In addition, in the e-commerce live broadcast management system, the image blocks enclosed by the edge pixels in the customized filtered image are used as reference image blocks. When the contour of a reference image block in the customized filtered image matches the baseline contour of the private part of the human body, an object recognition instruction is issued, including: the baseline contour of the private part of the human body is the geometric shape of the edge curve of the baseline pattern of the private part of the human body, and the number of private parts of the human body is multiple.

[0017] According to various embodiments of the present invention, the present invention has at least the following three main inventive features: Firstly, the image signal extraction operation is performed on the live broadcast room where the live broadcast user reports that the private parts of the human body are leaked, so as to randomly extract a single frame of live broadcast screen from multiple frames of live broadcast screen corresponding to multiple consecutive moments as the monitoring screen for targeted intelligent analysis of the private parts of the human body, and when the analysis result determines that the private parts of the human body are present, the live broadcast screen is directional coded, thereby improving the real-time performance of the shielding processing of the private parts of the human body in the live broadcast screen; Secondly: obtaining the brightness values ​​corresponding to the pixels in the received customized filtered image, taking each pixel in the customized filtered image as the center pixel, determining the brightness gradient of the center pixel based on the brightness values ​​of the center pixel and the pixels around it, and determining the center pixel as an edge pixel when the brightness gradient of the center pixel exceeds a limit, so as to obtain the edge pixels in the customized filtered image; Again: each image block enclosed by each edge pixel point in the customized filtered image is used as each reference image block, and when the contour of a certain reference image block in the customized filtered image matches the reference contour of the private part of the human body, an object recognition instruction is issued.

[0018] The e-commerce live broadcast management system of the present invention can solve the technical problems in the prior art that the coding of private parts of the human body in the live broadcast screen is not timely enough and the amount of data to be processed is large. It can perform image signal extraction operations on the live broadcast room reported by the live broadcast users as having leaked private parts of the human body, and perform targeted intelligent analysis on the existence of private parts of the human body for the randomly selected single-frame live broadcast screen as the monitoring screen, and perform targeted coding processing on the live broadcast screen when the analysis result determines that private parts of the human body exist, thereby reducing the processing calculation amount as much as possible and improving the real-time performance of the shielding processing of private parts of the human body in the live broadcast screen.

[0019] It is to be understood that, although the present invention has been disclosed as a preferred embodiment, the above embodiment is not intended to limit the present invention. For any technician familiar with the art, without departing from the scope of the technical solution of the present invention, the technical content disclosed above can be used to make many possible changes and modifications to the technical solution of the present invention, or modified into equivalent embodiments of equivalent changes. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of protection of the technical solution of the present invention.

Claims

1. An e-commerce live broadcast management system, characterized in that: The system comprises: The signal extraction mechanism is used to extract image signals from live broadcast rooms reported by live broadcast users to reveal private parts of human body, so as to randomly extract a single frame of live broadcast pictures from multiple frames of live broadcast pictures corresponding to multiple consecutive moments as monitoring picture output; An affine transformation device, connected to the signal extraction mechanism, for performing affine transformation processing on the received monitoring picture to obtain and output a corresponding affine transformation image; an edge sharpening device, connected to the affine transformation device, for performing edge sharpening processing on the received affine transformation image to obtain and output a corresponding edge sharpened image; a customized filtering device, connected to the edge sharpening device, for performing a salt and pepper noise removal action on the received edge sharpened image to obtain and output a customized filtered image, wherein the salt and pepper noise removal action is based on a median filtering algorithm; A targeted identification component is used to obtain the brightness values ​​corresponding to the pixels in the received customized filtered image, take each pixel in the customized filtered image as a central pixel, determine the brightness gradient of the central pixel based on the brightness values ​​of the central pixel and the pixels around it, and when the brightness gradient of the central pixel exceeds a limit, judge the central pixel as an edge pixel to obtain the edge pixels in the customized filtered image, and use the image blocks surrounded by the edge pixels in the customized filtered image as reference image blocks. When the contour of a reference image block in the customized filtered image matches the reference contour of the private part of the human body, an object identification instruction is issued; The coding processing component is connected to the targeted identification component and is used for coding the reference image block existing in the customized filtering image when receiving the object identification instruction.

2. The e-commerce live broadcast management system according to claim 1, characterized in that: The targeted identification component is also used to issue an object unidentified instruction when the contours of each reference image block in the customized filtered image do not match the reference contours of the private parts of the human body; The coding processing component is further configured to not perform coding processing on the reference image block existing in the customized filtered image when receiving the object unrecognized instruction.

3. The e-commerce live broadcast management system according to claim 2, characterized in that: The system further comprises: An optimization processing mechanism is connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, respectively, and is used to perform signal optimization processing on the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, so as to obtain respective optimized processed images corresponding to the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component; Among them, signal optimization processing is performed on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, including: the first step of optimization processing is to perform edge sharpening operation on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively.

4. The e-commerce live broadcast management system according to claim 3, characterized in that: The signal optimization processing is performed on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, and the second step of optimization processing also includes performing content enhancement operations on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively.

5. The e-commerce live broadcast management system according to claim 4, characterized in that: The signal optimization processing is respectively performed on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component. The third step of optimization processing also includes filtering processing operations on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component.

6. The e-commerce live broadcast management system according to claim 5, characterized in that: The signal optimization processing is performed on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively to obtain the optimized processed images corresponding to the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively, and also includes: the fourth step of optimization processing is to perform interpolation processing operations on the output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component respectively.

7. The e-commerce live broadcast management system according to any one of claims 3 to 6, characterized in that: The system further comprises: The data display mechanism is respectively connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, and is used to perform real-time display of each optimized processed image corresponding to the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component.

8. The e-commerce live broadcast management system according to claim 7, characterized in that: The data display mechanism is respectively connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, and is used to perform real-time display of each optimized processed image corresponding to the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, including: the data display mechanism is a liquid crystal display device.

9. The e-commerce live broadcast management system according to claim 7, characterized in that: The data display mechanism is respectively connected to the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, and is used to perform real-time display of each optimized processed image corresponding to the respective output image signals of the edge sharpening device, the customized filtering device, the targeted identification component and the coding processing component, including: the data display mechanism is a split-screen display device.