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

Infrared Sample Target Detection Method Based on Cyclic Consistent Adversarial Network

A target detection and network technology, applied in the field of image processing, can solve the problems of not improving the quality of target detection information, limited influence, and insufficient robustness of target detection network learning feature representation

Active Publication Date: 2022-03-15
TIANJIN UNIV
View PDF6 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, there are limitations in data expansion. It only expands the scale of the data set and does not improve the quality of target detection information, which will make the feature representation learned by the target detection network not robust enough.
In infrared small sample target detection, this limitation will have a greater impact

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
  • Infrared Sample Target Detection Method Based on Cyclic Consistent Adversarial Network
  • Infrared Sample Target Detection Method Based on Cyclic Consistent Adversarial Network
  • Infrared Sample Target Detection Method Based on Cyclic Consistent Adversarial Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] The proposed method mainly includes: enhancement of infrared image details of the original data set, modification of image generation network, training of image generation network, yolov3 image detection and other steps. figure 1 A block diagram of the proposed method is given. include:

[0039] 1. Raw data set infrared image detail enhancement

[0040] This patent selects the FLIR ADAS data set as the original data set. The data images contained in the data set are infrared images and visible light images of non-corresponding scenes, and 1366 infrared images and 1257 visible light images are selected as training images. The details of the infrared image are enhanced, the specific method is as follows:

[0041] For the original infrared image, select a local window w with pixel k as the center and r as the radius k , using equations (1) and (2) to calculate a k , b k :

[0042]

[0043]

[0044] Denote the input raw infrared image by I, For the input origi...

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 relates to a method for detecting an infrared sample target based on a cyclic consistent confrontation network, comprising the following steps: 1) enhancing the details of the infrared image of the original data set; modifying the image generation network; training the image generation network, using the data set enhanced by the infrared image details The adjusted image generation network is trained, the training data includes visible light images and infrared images processed by image detail enhancement; the final image generation network model is obtained; applied to the target detection network, the visible light image obtained through the image generation network model is input The object detection model performs object detection.

Description

technical field [0001] The invention belongs to the field of image processing and relates to an infrared small sample target detection method. Background technique [0002] The human visual system can only receive spectral information of visible light. Experts and scholars have proposed many advanced computer vision target detection methods in the field of visible light images. However, visible light images are easily affected by environmental factors such as weather and light intensity, and it is difficult to perform target detection tasks in certain environments. Infrared images represent the energy or temperature information of the detection target in the current scene by receiving thermal imaging of the target and background radiation, and the information is rarely affected by environmental factors. Therefore, infrared image target detection has become an important research direction. [0003] With the development of deep learning technology, infrared image target dete...

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): G06T7/00G06T5/00G06N3/04G06N3/08
CPCG06T7/0002G06T2207/10048G06T2207/20081G06N3/045
Inventor 杨嘉琛李爱云
Owner TIANJIN 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