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

Deep learning-fused pedestrian posture multi-feature intelligent identification method

A deep learning, multi-feature technology, applied in character and pattern recognition, instruments, biological neural network models, etc., can solve the problems of simple pedestrian tracking processing results and low accuracy

Active Publication Date: 2018-11-20
CENT SOUTH UNIV
View PDF14 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention provides a multi-feature intelligent identification method for pedestrian postures that integrates deep learning, and its purpose is to overcome the situation that the results of pedestrian tracking processing are simple and the accuracy is not high in the surveillance video of the prior art

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
  • Deep learning-fused pedestrian posture multi-feature intelligent identification method
  • Deep learning-fused pedestrian posture multi-feature intelligent identification method
  • Deep learning-fused pedestrian posture multi-feature intelligent identification method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0075] The present invention will be further described below in conjunction with accompanying drawings and examples.

[0076] Such as figure 1 As shown, a multi-feature intelligent recognition method for pedestrian postures fused with deep learning includes the following steps:

[0077] Step 1: Construct a pedestrian sample image database;

[0078] The pedestrian sample image database is to extract continuous pedestrian image frames from the intersection monitoring video, and obtain three types of image groups;

[0079] The three types of image groups are respectively negative samples without pedestrians, multi-person samples containing multiple pedestrians, and single-pedestrian samples containing only the same pedestrian, and each type of image group includes at least 300 frames of images;

[0080] Step 2: Preprocess the pedestrian image frames in the pedestrian sample database, and set the pedestrian detection frame, pedestrian target identification and pedestrian locatio...

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-fused pedestrian posture multi-feature intelligent identification method. The method extracts a pedestrian detection frame and a pedestrian position label vector, uses the pedestrian position label vector as a pedestrian trajectory point, and can accurately track all pedestrians. Compared with the prior art, the detection is accurate. The neural network based deep learning method is adopted in the pedestrian detection, and the pedestrian can be quickly and effectively detected and marked, which can meet the requirements for the instant identification ofthe emergency in an actual traffic environment. The method is also suitable for complex environments such as intelligent factories, laboratories, robots, etc.; the re-identification rate is high, andthe neural network automatically extracts high-level abstract features from tracking targets to achieve efficient and rapid matching and re-identification of the tracking targets; and an optimizationalgorithm is used to adjust and optimize network parameters, the accuracy of pedestrian identification is improved, and high robustness is achieved.

Description

technical field [0001] The invention belongs to the field of artificial intelligence, and in particular relates to a multi-feature intelligent identification method of pedestrian postures fused with deep learning. Background technique [0002] With the rapid development of science and technology, pedestrian detection technology using computer vision related technologies has been widely used in all aspects of life, such as intelligent trains, vehicle automatic driving and other fields. Traffic safety is an eternal topic. In vehicle collision accidents, collisions between vehicles and pedestrians also account for a large proportion. Nowadays, traditional safety technologies such as seat belts and airbags have been widely used, but these are passive protection methods. People hope to study the active protection safety system of vehicles, and the accurate identification and tracking of pedestrians is the focus of research. [0003] At present, the mostly used pedestrian track...

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): G06K9/00G06N3/04
CPCG06V40/20G06N3/045
Inventor 刘辉李燕飞黄家豪韩宇阳
Owner CENT SOUTH UNIV
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