Unlock instant, AI-driven research and patent intelligence for your innovation.

Target identification method, target identification apparatus, terminal device and target identification system

A technology of target recognition and target category, applied in the field of target recognition, it can solve the problem that the data type cannot simulate the scene, etc., and achieve the effect of high accuracy.

Inactive Publication Date: 2017-10-31
HITACHI LTD
View PDF11 Cites 23 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Moreover, the training samples for online learning only come from the first to K frames of pictures, and the data type cannot simulate more scenes

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
  • Target identification method, target identification apparatus, terminal device and target identification system
  • Target identification method, target identification apparatus, terminal device and target identification system
  • Target identification method, target identification apparatus, terminal device and target identification system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0061] Various embodiments according to the present invention will be described below with reference to the accompanying drawings.

[0062] figure 1 A schematic diagram showing the configuration of the object recognition system 1 according to the embodiment of the present invention.

[0063] like figure 1 As shown, the object recognition system 1 includes a remote server 10 and a plurality of terminal devices 30 . Each terminal device 30 is connected to the remote server 10 via the network 20 so as to be able to exchange data with each other.

[0064] Each terminal device 30 includes an object recognition device 300 . The target recognition device 300 of each terminal device 30 uses the stored first classifier to perform target recognition on the image data collected in real time, stores the image data satisfying the predetermined storage conditions as training samples, and uses all the stored training samples The third classifier obtained through deep learning processing ...

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 provides an online automatic deep learning-based target identification method and apparatus, a terminal device and a target identification system. The method comprises the steps of collecting image data in real time; by utilizing a stored first classifier, performing target identification on the collected image data to generate image data with a category label, wherein the category label comprises a target category and a target posterior probability; when the target posterior probability meets a predetermined storage condition, performing storage as a training sample according to the target category; and when a predetermined starting condition is met, starting online deep learning processing to obtain a third classifier, and updating the stored first classifier by using the obtained third classifier. By adopting the method, high-accuracy target identification can be realized.

Description

technical field [0001] The present invention relates to target recognition technology, in particular to a target recognition method based on online automatic deep learning, a target recognition device, a terminal device and a target recognition system. Background technique [0002] Object recognition technology has broad application prospects in video surveillance, robotics, intelligent transportation and other fields. However, due to the calculation and analysis of a large amount of data involved in target recognition, and the interference of environmental factors such as light angle of view, traditional recognition algorithms cannot extract the preferred features of the image, resulting in a limited recognition rate. [0003] One method of object recognition can adopt traditional offline learning methods. However, in traditional offline learning methods, the classification model trained with an offline classifier is only suitable for a specific target in a specific enviro...

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): G06K9/62
CPCG06F18/243
Inventor 戴依若
Owner HITACHI LTD
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com