A Single Target Tracking Method Based on Siamese Network

A single-target, network technology, applied in the field of target tracking, can solve problems such as tracking failure, lack of template update, poor tracking effect, etc., achieve obvious effects, reduce training costs and computing overhead, and improve tracking processing efficiency

Active Publication Date: 2022-05-03
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF11 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] (1) Lack of necessary template updates;
[0010] (2) Since the window width remains unchanged during the tracking process, when the target scale changes, the tracking will fail;
[0011] (3) When the target speed is fast, the tracking effect is not good;
[0012] (4) The histogram feature is slightly deficient in the description of the target color feature, and lacks spatial information

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 Single Target Tracking Method Based on Siamese Network
  • A Single Target Tracking Method Based on Siamese Network
  • A Single Target Tracking Method Based on Siamese Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0111] The COCO dataset is a 640×480 RGB image, and 100 images were randomly selected from the dataset as training data. The selected image data are as follows: Figure 4 Shown; Then the screened image is cropped to 511×511 and sent to the training network.

[0112] Experiment on the COCO dataset, train with the improved ResNet50 as the skeleton network, and set different parameters and network structures. Using 0TB2015 as the evaluation data set, the specific experimental results are as follows Figure 5 As shown, among them, Tracnker name represents the model parameter weights of different batches of training, Success represents the success rate of tracking, and Prectision represents the accuracy of tracking.

[0113] In this embodiment, after the Siamese feature extraction sub-network is pre-trained on ImageNet, the network is trained on the training set of the COCO dataset, and the size of the training set exceeds 20GB. In training and testing, a single-scale image with ...

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 single target tracking method based on a Siamese network, belonging to the technical field of target tracking. The present invention first constructs the neural network part of the Siamese network and trains the weight of the Siamese convolutional neural network. In the training process, the neural network model based on the embedded loss function is carried out. At the same time, the features of each layer are fused, and the stochastic gradient descent algorithm is used for loss. Optimization, and then obtain the results of classification and regression through RPN, and finally track the subsequent frames for the results of classification and regression. The invention can better detect and locate the tracking target, can effectively integrate the image detection method into the target tracking, and uses the image frame instead of the video, thereby reducing the training cost and calculation overhead. Thereby, the tracking processing efficiency is improved, and the effect of distinguishing similar objects is more obvious.

Description

technical field [0001] The invention belongs to the technical field of target tracking, in particular to a single target tracking technology based on Siamese network prediction technology. Background technique [0002] With the rapid development of hardware, software and artificial intelligence, target tracking has become one of the hot spots in the field of computer vision research and has been widely used. The tracking and focusing of cameras and the automatic target tracking of drones all require target tracking technology. In addition, there are specific object tracking, such as human body tracking, vehicle tracking in traffic monitoring systems, face tracking and gesture tracking in intelligent interaction systems. Simply put, target tracking is to establish the positional relationship of the object to be tracked in a continuous video sequence, and obtain the complete motion trajectory of the object. Given the coordinate position of the target in the first frame of th...

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): G06V20/40G06V10/26G06N3/04G06N3/08
CPCG06N3/08G06V20/48G06V20/46G06V20/41G06V10/267G06N3/045
Inventor 饶云波程奕茗郭毅薛俊民
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap