remote sensing image rapid target detection method based on a deep Hash auxiliary network

An auxiliary network and remote sensing image technology, applied in the field of computer vision, can solve problems such as waste of computing resources, lower detection efficiency, and sparse target distribution, and achieve the effects of increasing correlation, improving detection efficiency, and simplifying calculations

Active Publication Date: 2019-06-21
XIDIAN UNIV
View PDF3 Cites 41 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Wide-format remote sensing images generally refer to remote sensing images with a size of more than 10,000×10,000. The target distribution in the image is relatively sparse, that is, most areas do not contain targets. If most non-target background areas are processed uniformly without discrimination, it will inevitably cause Waste of computing resources, reducing detection efficiency

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
  • remote sensing image rapid target detection method based on a deep Hash auxiliary network
  • remote sensing image rapid target detection method based on a deep Hash auxiliary network
  • remote sensing image rapid target detection method based on a deep Hash auxiliary network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0026]With the rapid development of satellite sensor technology, the acquisition of high-spatial-resolution wide-range remote sensing images has become more and more convenient, and the rapid target detection technology for high-resolution wide-range remote sensing images has become a popular research direction in the field of remote sensing images , this technology has been widely used in multiple application scenarios such as rapid identification and positioning of military targets and crop detection in civilian use, and people have more and more application requirements for this technology in both military and civilian fields.

[0027] Although the target detection technology of remote sensing images is constantly improving in terms of hardware and detection performance, the existing methods are still aimed at small images that are similar in size to natural images. Carry out carpet-type indiscriminate detection in each area of ​​the image, but the distribution of targets in...

Embodiment 2

[0050] The remote sensing image fast target detection method based on deep hashing auxiliary network is the same as that in embodiment 1,

[0051] The deep hash assisted network construction described in step (2) includes the following steps:

[0052] (2a) Use the low-level structure of the pre-trained model deep residual network (ResNet101) trained on the large-scale image classification dataset ImageNet as the basic network extraction feature of the deep hashing auxiliary network to obtain the basic network feature map, and the image block sample s j is the input to the deep hashing auxiliary network.

[0053](2b) The deep residual network consists of the first convolutional unit and four convolutional blocks. The first convolution unit and the first three convolution blocks are defined as the low-level structure, and the fourth convolution block is defined as the high-level feature extraction network module. The specific structure of the network is as follows: the first c...

Embodiment 3

[0065] The method for fast target detection in remote sensing images based on the deep hash-assisted network is the same as in embodiment 1-2, and the layer structure of the hash-assisted branch network in the step (2) of the present invention is as follows: input layer→multiplexing convolution layer→adaptive Pooling layer → fully connected layer → activation function layer → binary classifier layer → output layer.

[0066] see figure 2 , the hash-assisted branch network in the present invention is built on the output end of the basic network feature map, and its input layer is the input of the hash-assisted branch network, that is, the basic network feature map described in step (2). The multiplexing convolutional layer shares parameters in the hash-assisted branch network and the candidate region generation network, that is, a feature multiplexing mechanism is introduced in different tasks, so that the features extracted by the multiplexing convolutional layer contain both ...

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 remote sensing image rapid target detection method based on a deep Hash auxiliary network, and solves the problem that an existing target detection framework cannot carry outefficient detection on a background area which does not have a target and occupies a large area in a wide remote sensing image. The method comprises the following implementation steps: generating anexperimental training image block set and a test image block set; constructing a remote sensing image rapid target detection model based on a deep Hash auxiliary network; training a remote sensing image rapid target detection model based on a deep Hash auxiliary network; and using the trained deep Hash auxiliary network to obtain a test result of the test image block set, and counting the runningtime. According to the method, the Hash auxiliary branch network module is introduced, redundant calculation in target detection is greatly reduced by quickly filtering most non-target background areas in the wide remote sensing image, the detection time of the non-target background areas is greatly shortened, and the detection efficiency is greatly improved. The method is used for target detection of wide remote sensing images.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and further relates to the detection of remote sensing image targets, in particular to a remote sensing image rapid target detection method based on a deep hash auxiliary network. Object detection for high-resolution remote sensing images. Background technique [0002] With the advancement of sensor technology, researchers can now easily obtain images with high spatial resolution or spectral resolution. The rapid target detection technology for high-resolution images is becoming a hot research field in the field of images. It is used in both military and civilian applications. Can be widely used. [0003] Deep learning has achieved great success in the field of image object detection. Girshick et al. proposed a two-stage target detection framework in their paper "Rich feature hierarchies for accurate object detection and semantic segmentation" (IEEE conference on computer vision and pat...

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): G06K9/00G06K9/46G06N3/04G06N3/08
Inventor 杨淑媛王喆冯志玺胡滔徐光颖余亚萍马文萍尚荣华王敏焦李成
Owner XIDIAN UNIV
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