Method for removing video watermarks based on deep learning

A video watermarking and deep learning technology, applied in the field of image processing, can solve problems such as poor performance, time-consuming and labor-intensive video processing, and traces of watermarked areas, so as to improve the ability to remove watermarks, improve generalization capabilities, and reduce large batches. The effect of production demand

Inactive Publication Date: 2020-11-10
GUILIN UNIV OF ELECTRONIC TECH
View PDF12 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Usually, the watermark processing software on the market removes a single watermark image. This method is not only time-consuming and labor-intensive for video processing, but also has obvious traces in the watermark area, and the effect is not good.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method for removing video watermarks based on deep learning
  • Method for removing video watermarks based on deep learning
  • Method for removing video watermarks based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0024] refer to figure 1 , a method for removing video watermarks based on deep learning, comprising the following steps:

[0025] 1) Video processing: use the video processing opencv tool to convert the video into pictures, and divide the pictures into pictures with watermarks and pictures without watermarks;

[0026] 2) Data enhancement: Data enhancement can increase the diversity of data, enhance the robustness of simulation, reduce the sensitivity of the model to parameters, and improve the generalization ability of the model. The use of mosaic data enhancement can improve the ability to remove watermarks. The process : Randomly add watermark noise, read 4 pictures each time, make the same transformation of flipping, zooming, and rotating the pictures with watermark and pictures without watermark at the same time, and arrange them according to the four directions, and combine the pictures. Combining frames to synthesize a picture, so that the background of the detected ob...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a method for removing video watermarks based on deep learning, which is characterized by comprising the following steps: 1) video processing; 2) data enhancement; 3) model training; and 4) realizing watermark elimination. The method has the advantages that the efficiency of processing massive video watermarks can be improved, the video watermarks can be rapidly removed in batches, no trace is left for watermark removal, and frames are not lost in original image quality after the watermarks are removed.

Description

technical field [0001] The invention relates to image processing technology, in particular to a method for removing video watermarks based on deep learning. Background technique [0002] With the rapid development of network technology, various videos are spread on the Internet. In order to protect the copyright in videos, it is a common practice for enterprises to superimpose watermarks in videos. In many cases, users need to use videos without watermarks to facilitate their own business. Usually, the watermark processing software on the market removes a single watermark image. This method is not only time-consuming and labor-intensive for video processing, but also has obvious traces in the watermark area, and the effect is not good. Therefore, how to efficiently process the watermark in the video has become an urgent problem to be solved in the industry. Contents of the invention [0003] The purpose of the present invention is to provide a method for removing video ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): H04N21/8358G06T1/00G06N3/08G06N3/04
CPCH04N21/8358G06T1/0021G06N3/08G06T2201/0065G06N3/045
Inventor 覃琴颜靖柯王鑫
Owner GUILIN UNIV OF ELECTRONIC TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap