Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Image fuzzy region automatic detection method and system

A fuzzy area, automatic detection technology, applied in the field of image processing, can solve the problems of reducing labor costs, manual labeling data dependence, etc., to achieve the effect of improving the degree of automation, reducing time and economic costs, and optimizing algorithm performance

Active Publication Date: 2017-07-04
BEIJING QIYI CENTURY SCI & TECH CO LTD
View PDF7 Cites 36 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At the same time, it can obtain a large amount of data with a small amount of manual labeling, train the deep learning model, solve the problem of dependence on manual labeling data in the process of deep learning, and greatly reduce labor costs

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
  • Image fuzzy region automatic detection method and system
  • Image fuzzy region automatic detection method and system
  • Image fuzzy region automatic detection method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0050] It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

[0051] The present invention provides a method for automatic detection of image blurred areas, referring to figure 1 , in the first embodiment, the automatic detection method of the blurred image area includes:

[0052] Step S10, extracting clear image samples and blurred image samples from a preset image database;

[0053]Step S20, using the clear image samples, blur image samples and preset network structure to train the parameters of the deep convolutional neural network model to obtain network models of clear image samples and blur image samples;

[0054] In the present invention, a certain number of images can be prepared in advance, and the images contain at least some clear areas, and these images can be obtained from many open-source image databases. For each image in the image library, a certain amount of r...

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 an image fuzzy region automatic detection method and system. The method comprises the following steps of extracting a clear image sample and a fuzzy image sample from a preset image database, training parameters of a deep convolutional neural network model by utilizing the clear image sample, the fuzzy image sample and a preset network structure, and obtaining network models of the clear image sample and the fuzzy image sample; inputting all small image blocks to the trained model, and performing prediction of the deep convolutional neural network models of the small image blocks in sequence according to the same model structures and the trained parameters, until clear confidence degrees and fuzzy confidence degrees of all the small image blocks are finally obtained; and performing operation on the clear confidence degrees and the fuzzy confidence degrees of all the small image blocks to obtain a definition score of a whole image. According to the method and the system, the time and economic costs of manual tagging can be reduced, the automation degree can be improved, and the network performance can be ensured.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an automatic detection method and system for image blurred areas. Background technique [0002] Clarity is an evaluation index that is crucial to measure image quality. For image users or producers, it is expected to use or produce images with clear subjects to express the meaning they want to convey. For example, for image producers, when users take photos or videos, although the current focusing technology and anti-shake technology can help users improve the quality of images to a certain extent, these technologies cannot guarantee that all captured images and videos are perfect. clear. Due to the failure of focusing, the strong movement of the object being photographed, or the hand shake of the photographer during the shooting process, the user may obtain blurred photos or videos without knowing it, and lose the opportunity to re-shoot. If there is a technology for ...

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
IPC IPC(8): G06T7/00
CPCG06T7/0002G06T2207/10004G06T2207/10016G06T2207/20021G06T2207/20076G06T2207/20081G06T2207/20084G06T2207/30168
Inventor 刘楠
Owner BEIJING QIYI CENTURY SCI & TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com