Multi-source remote sensing image fusion target comprehensive detection method
A technology of remote sensing image fusion and target detection, which is applied in instrument, computing, character and pattern recognition, etc., to achieve high tolerance, improve accuracy and reliability.
Active Publication Date: 2022-01-21
10TH RES INST OF CETC
View PDF10 Cites 5 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0011] The present invention aims at optical remote sensing images with complex backgrounds and the deficiencies of existing technologies, and provides a smarter and more accurate detection method that can improve the detection effect of complex scenes. A comprehensive target detection method based on multi-source image fusion with high efficiency and reliability
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0022] refer to figure 1. According to the present invention, the visible light Image_VIS and the infrared image Image_IR that are paired in the same area are obtained by using the visible light and infrared sensors mounted on the aircraft and the satellite earth observation platform. The visible light and infrared earth observation image data are paired one by one, and the image resolution is unified and registered, and the targets on the remote sensing image are marked at the same time, and the visible light, Infrared, visible light-infrared images pair three remote sensing single-source and multi-source target detection data sets, and divide training set, verification set and test set; based on deep neural network, according to visible light, infrared single-source and multi-source target detection data sets The training set samples were used to build visible light and infrared single-source detection network models and multi-source target detection network models based on...
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
Login to View More Abstract
The multi-source remote sensing image fusion target comprehensive detection method disclosed by the invention is high in detection accuracy and reliability. The method is realized through the following technical scheme: based on target comprehensive detection, inputting paired visible light and infrared earth observation image data, constructing visible light and infrared single-source and multi-source target detection data sets, and according to a training sample of each data set, constructing visible light, infrared single source and image feature level fused multi-source target detection network models respectively; the multi-source target detection network model extracting complementary features of different sources for interactive learning to obtain multi-scale multi-source fusion features, and predicting the position and category of a target in the image based on a multi-scale detection network; and inputting the single-source test data set and the multi-source test data set into a single-source target detection model and a multi-source target detection model respectively, outputting corresponding single-source and multi-source target detection results by the models, and outputting a target comprehensive detection result in a decision-level fusion manner according to a target-associated detection result.
Description
technical field [0001] The invention relates to a target detection method widely used in remote sensing, computer vision, image processing and other fields, especially in remote sensing images acquired by platforms such as low-altitude drones, aviation planes and satellites. Background technique [0002] Object detection is a basic and important key technology in the field of computer vision. It has received a lot of attention in recent years and has been widely used in intelligent video surveillance, intelligent transportation, robot visual navigation, medical diagnosis, automatic driving and other industries. . The purpose of target detection is to separate the target of interest from the background in a given image, and determine the location and category of the target of interest; in recent years, target detection based on remote sensing images has gradually become an important technology In the field of research, it is often used in the rapid detection and tracking 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
Login to View More IPC IPC(8): G06V10/80G06V10/774G06V20/40G06K9/62
CPCG06F18/253G06F18/256G06F18/214
Inventor 李雪向涛赖作镁
Owner 10TH RES INST OF CETC