Target detection method combined with lightweight network

A target detection and lightweight technology, applied in the field of target detection and deep learning, can solve the problems of the difficulty of drone detection and supervision, and achieve the effect of improving detection performance, fast running speed and high accuracy

Pending Publication Date: 2021-06-22
INST OF OPTICS & ELECTRONICS - CHINESE ACAD OF SCI
View PDF9 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide a target detection method combined with a lightweight network to solve the problem that the detection and supervision of unmanned aerial vehicles are very difficult. Experiments have proved that the method propos

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 detection method combined with lightweight network
  • Target detection method combined with lightweight network
  • Target detection method combined with lightweight network

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0031] A detailed description of the specific embodiments of the present invention will be described below in conjunction with the accompanying drawings. However, the following examples are limited to the appended claims, and the scope of the invention should be limited by the claims of the claims, and all the contents of the claims of the invention will be implemented by those skilled in the art.

[0032] figure 1 For the present invention, three feature scale detects to five feature scales are detected: in order to better apply to small target detection scenes such as drones, the YOLOV3 algorithm that was detected by three feature scale. Extension is five feature scale to detect, and better performance in the detection scenario of small objectives;

[0033] figure 2 A comparison of the conventional convolution of the present invention with a comparison of GHOST MODULE: Compared to conventional convolution, the GHOST module generates more features through a series of linear tra...

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 a target detection method combined with a lightweight network, and aims to solve the problem of balancing the detection speed and precision of a small target such as UAV, realize the lightweight of a network model, and provide possibility for target detection on an embedded platform. According to the characteristics of small size, high flight speed and the like of an unmanned aerial vehicle, the invention provides a quick target detection algorithm improved in combination with a lightweight network. The YOLOv3 algorithm for detecting three feature scales is expanded to five feature scales for detection, and the detection performance of small target scenes such as unmanned aerial vehicles is improved. A lightweight feature extraction network is constructed in combination with a Ghost module in a lightweight network, and in order to further improve the detection performance of the network, a channel attention mechanism is added to suppress unfavorable information. According to the invention, a UAV data set of an urban background is generated for training. Experimental results show that the method provided by the invention can effectively improve the detection precision of the UAV under the complex city background and meet the real-time requirement.

Description

technical field [0001] The invention relates to the fields of deep learning and target detection, in particular to a target detection method combined with a lightweight network. Background technique [0002] With the development of science and technology, various kinds of UAVs have appeared one after another. Among them, civilian UAVs have developed rapidly and are widely used in various fields. The application of UAVs can greatly reduce the cost of high-altitude operations. Unique advantages, UAV technology brings convenience to people's life, but also brings many potential threats to the country and society. The supervision of drones has become the focus of social attention. UAVs have the characteristics of small size, fast flight speed, and high rotation. Traditional target detection methods are difficult to meet the requirements. The target detection method based on deep neural network shows its powerful detection performance. In recent years, the rapid development of ...

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): G06K9/00G06K9/46G06K9/62
CPCG06V20/176G06V10/40G06V2201/07G06F18/214
Inventor 毛耀李鸿杨锦辉彭锦锦胡钦涛刘超杜芸彦
Owner INST OF OPTICS & ELECTRONICS - CHINESE ACAD OF SCI
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