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

Unmanned aerial vehicle face recognition method based on super-resolution

A technology of super-resolution and recognition methods, applied in neural learning methods, character and pattern recognition, image analysis, etc., to achieve the effect of low power consumption and low latency

Pending Publication Date: 2022-06-10
HOHAI UNIV
View PDF11 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the traditional mobile face recognition is a ground maneuvering method, and there are still problems such as building occlusion, plane maneuvering limitation, and narrow field of view. Therefore, with the popularization and application of UAVs, high-speed Mobile tracking and monitoring technology has become a research hotspot

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
  • Unmanned aerial vehicle face recognition method based on super-resolution
  • Unmanned aerial vehicle face recognition method based on super-resolution
  • Unmanned aerial vehicle face recognition method based on super-resolution

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0025] The model uses the bottleneck of MobileNetV2 as the main module to build the network, and the bottleneck in MobileFaceNet is smaller than that of MobileNetV2. In addition, fast downsampling is used at the beginning of the network, early dimensionality reduction is used in the last few convolutional layers, and a linear convolutional layer is added after the linear global depth convolutional layer as the feature output. Batch regularization is employed during training.

[0026] Add SENet to the MobileFaceNet network structure. After adding the specific position of the depthwise conv3×3 of the bottleneck, the MobileFaceNet model based on the attention mechanism is obtained. After the UAV face image is input, it is first processed by image preprocessing and adjusted to a size of 112×112. After that, the image goes through conv3×3 and depthwise conv3×3. In the bottleneck part, SENet is introduced, and in conv1 ×1, Linear GDConv7×7 and Linear GDConv7×7 and then output the ...

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 unmanned aerial vehicle face recognition method based on super-resolution, and the method comprises the following steps: 1, shooting videos of people in various forms under the view angle of an unmanned aerial vehicle, obtaining a face bounding box through labeling, carrying out the clustering of a target face, and building an unmanned aerial vehicle face data set; step 2, constructing a residual dense module, combining the super-resolution model SRGAN with the residual dense module, and training the super-resolution model; step 3, realizing processing of the unmanned aerial vehicle face picture from low resolution to high resolution; and 4, the unmanned aerial vehicle face data set is recognized through the lightweight face recognition model based on the attention mechanism, the lightweight model parameter quantity is small, and the recognition accuracy is improved again.

Description

technical field [0001] The invention relates to a UAV face recognition method, in particular to a UAV face recognition method based on super-resolution, and belongs to the field of artificial intelligence recognition technology. Background technique [0002] At present, the application of drones is very common, and image extraction is the most intuitive and effective identification method. For the positioning and identification technology of UAVs, there are many research plans, which also involve research in different fields. When tracking targets in open outdoor places, the fixed face recognition mode has natural shortcomings such as monitoring blind spots and inability to maneuver. For this reason, people adopt mobile face recognition to overcome above-mentioned defect. However, the traditional mobile face recognition is a ground maneuvering method, and there are still problems such as building occlusion, plane maneuvering limitation, and narrow field of view. Therefore,...

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 Applications(China)
IPC IPC(8): G06V40/16G06V20/17G06V10/74G06V10/774G06V10/82G06T3/40G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06T3/4053G06T2207/10016G06T2207/20081G06T2207/20084G06T2207/30201G06N3/045G06F18/22G06F18/214
Inventor 刘凡庄芸王菲许峰
Owner HOHAI 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