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

An Adaptive Particle Filter Target Tracking Method Based on Deep Learning

A particle filtering and target tracking technology, applied in neural learning methods, instruments, biological neural network models, etc., can solve problems such as complex backgrounds, and achieve the effects of enhancing expression ability, improving adaptability, and improving robustness

Active Publication Date: 2020-04-21
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF7 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide an adaptive particle filter target tracking method based on deep learning, which can solve the problems of fast movement and complex background in existing target tracking, so as to realize continuous and robust tracking

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
  • An Adaptive Particle Filter Target Tracking Method Based on Deep Learning
  • An Adaptive Particle Filter Target Tracking Method Based on Deep Learning
  • An Adaptive Particle Filter Target Tracking Method Based on Deep Learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] like figure 1 As shown, an adaptive particle filter target tracking method based on deep learning includes the following steps:

[0022] The present invention provides an adaptive particle filter target tracking method based on deep learning, the design principle of which is as follows: construct a shallow deep learning network, and use the SGD method to train a stable network model offline; then apply it to the particle filter tracking framework; The output features of the deep learning network and the color histogram features are fused as the particle observation model; then the first-order and second-order motion information is introduced into the particle dynamic transfer equation to estimate the particle state; the target position is optimally estimated according to the particle state and observation model; finally, according to The target state change degree adopts the threshold method to update the target template.

[0023] The detailed steps of the adaptive par...

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 adaptive particle filter target tracking method based on deep learning, comprising the following steps: (1) establishing a shallow deep learning network model, and using the SGD algorithm to train the model offline; (2) applying the trained network Based on the particle filter tracking framework, for each particle, use depth features and color appearance features to solve the observation model; solve the state transition model according to the first-order and second-order motion information; (3) calculate the weight parameters of each particle according to the observation model , determine the current target position by weighted average; (4) Calculate the current state of the target, determine the degree of state change, and update the observation model in real time. The present invention provides an adaptive particle filter target tracking method based on deep learning, which can solve the problems of fast movement and complex background in existing target tracking, thereby realizing continuous and robust tracking.

Description

technical field [0001] The invention relates to the technical field of particle filter target tracking, in particular to an adaptive particle filter target tracking method based on deep learning. Background technique [0002] Video object tracking has become a cutting-edge research hotspot in the field of computer vision. Its main task is to obtain the position and motion information of the target of interest in the video sequence, and provide the basis for further semantic analysis. The research on video object tracking is widely used in intelligent video surveillance, human-computer interaction, medical diagnosis and other fields, and has strong practical value. [0003] Tracking algorithms can be divided into discriminative and generative. Discriminative methods model the tracking problem as a binary classification problem to distinguish objects from backgrounds. Generative methods search image regions with minimal reconstruction error by building object representation...

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): G06K9/62G06N3/08
CPCG06N3/08G06V10/751G06F18/2411
Inventor 钱小燕韩磊王跃东张艳琳张代浩
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
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