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

A Deep Learning-Based Image Target Picking Method

A deep learning and image technology, applied in the field of image target extraction based on deep learning, can solve problems such as lack of data sets, and achieve the effect of high practical value and correct identification

Active Publication Date: 2021-07-27
EAST CHINA NORMAL UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] An image target extraction method based on deep learning, which is characterized by the establishment of a mixed data set of natural images and synthetic images, which solves the problem of lack of data sets in the problem of image target extraction

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
  • A Deep Learning-Based Image Target Picking Method
  • A Deep Learning-Based Image Target Picking Method
  • A Deep Learning-Based Image Target Picking Method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0049] The present invention will be further described below in conjunction with the accompanying drawings.

[0050] This embodiment is implemented under the Windows 10 64-bit operating system on the PC, and its hardware configuration is CPU i5-6500, memory 16G, GPU NVIDIA GeForce GTX 1060 6G. Deep learning library Keras 2.0.8, which uses Tensorflow1.3.0 as the backend. Programming adopts Python language.

[0051] The method of image target extraction based on deep learning is characterized by: the establishment of a mixed dataset of natural images and synthetic images, which solves the problem of scarcity of datasets, and applies the features obtained from low-resolution learning to the foreground of high-resolution images In the learning process of the target, a learning mechanism combining low resolution and high resolution can be constructed to realize the extraction function of the image foreground target. The input image resolution of this method is N t ×N t , N t 22...

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 deep learning-based image target extraction method. In the foreground extraction strategy, a deep learning framework combining low resolution and high resolution is established, which solves the problem of uneven quality of data sets, and defines A reasonable network structure and loss function are established to ensure the quality and performance of the deep learning model. Through the establishment of a mixed data set of natural images and synthetic images, the problem of lack of data sets in image target extraction is solved. The invention avoids the limitation of the traditional tripartite image input, realizes the automatic extraction of the foreground object, and can obtain a finer image extraction result.

Description

technical field [0001] The invention relates to the technical field of image synthesis, in particular to an image target extraction method based on deep learning, which solves the problem of lack of data sets by establishing a mixed data set of natural images and synthetic images. Further, the features obtained by low-resolution learning are applied to the learning process of high-resolution image foreground objects, and a learning mechanism combining low-resolution and high-resolution is constructed to realize the extraction of image foreground objects. Background technique [0002] In recent years, with the continuous development of computer technology, in practical applications in e-commerce, medicine, entertainment and other fields, it is increasingly necessary to understand the real world through collected digital images and videos. The traditional image extraction method based on color sampling samples the pixels in the unknown area to be solved in the given foreground...

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 Patents(China)
IPC IPC(8): G06T7/12G06N3/04G06N3/08
CPCG06T7/12G06T2207/20081G06T2207/20084G06N3/08G06N3/048
Inventor 全红艳沈卓荟
Owner EAST CHINA NORMAL UNIV
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