SAR image iron tower target detection method based on deep learning

A target detection and deep learning technology, applied in the field of image processing, can solve problems such as poor detection performance

Inactive Publication Date: 2019-02-12
BEIHANG UNIV
View PDF1 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is to solve the problem that the target detection method for SAR image

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
  • SAR image iron tower target detection method based on deep learning
  • SAR image iron tower target detection method based on deep learning
  • SAR image iron tower target detection method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0078] Such as figure 1 As shown, an embodiment of the present invention provides a deep learning-based SAR image iron tower target detection method, including:

[0079] S1. Randomly extract multiple SAR images from the SAR data set, segment to obtain sample slices, set a sample label for each sample slice, and then construct a training sample set. The sample label includes the coordinate information and category information of the target in the sample slice.

[0080] Preferably, step S1 includes:

[0081] S1-1. Randomly extract 100 SAR images from the MiniSAR data set;

[0082] S1-2. Randomly segment each SAR image to obtain multiple sample slices;

[0083] S1-3. Add a sample label to each sample slice. The sample label includes the absolute path of the sample slice, the number of targets, the coordinates of the target frame (ie coordinate information) and the target category (ie category information); specifically, the sample label The composition can be:

[0084] n x t1 y t1 x b1 y ...

Embodiment 2

[0107] Such as figure 2 As shown, the second embodiment is basically the same as the first embodiment, and the similarities will not be repeated here. The differences are:

[0108] The input image size of the SSD model constructed in step S3 is 300×300 pixels.

[0109] figure 2 It is a schematic diagram of the SSD model in this embodiment, and the key convolutional layer is marked in the figure. Among them, the convolutional layer 4_3 represents the third convolutional layer in the fourth group of convolutional layers, that is, the 10th convolutional layer, the same below.

[0110] Such as figure 2 As shown, the specific structure of the 15-layer convolutional neural network constructed in step S3 includes:

[0111] The first group of convolutional layers: the first 1-2 convolutional layers, each layer uses 64 convolution kernels with a window size of 3×3, the window moving step is 1, the edge is filled with 1 pixel, and the output is 64 sizes 300×300 feature map;

[0112] The seco...

Embodiment 3

[0159] Such as image 3 , Figure 4a with Figure 4b As shown, the third embodiment is basically the same as the second embodiment, and the similarities will not be repeated here. The differences are:

[0160] The data to be tested used in this embodiment is a scene SAR image, such as image 3 Shown. The image size of the data map to be tested is 16384×8192 pixels, which contains a variety of artificial targets such as iron towers and buildings. The purpose is to detect and locate all types of iron tower targets in the data map to be tested.

[0161] In order to better observe the test results, such as Figure 4a with 4b As shown, the part of the target detection result is enlarged, and the part marked with the white frame in the figure is the part of the iron tower detected.

[0162] In this embodiment, by calculating the detection results of the entire data graph to be tested, it is concluded that the accuracy rate of the method provided by the present invention is 82.4%, the recal...

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 SAR image iron tower target detection method based on deep learning. The method comprises the following steps of randomly extracting a plurality of SAR images from a SAR data set, segmenting and obtaining sample slices, setting sample labels for each sample slice and constructing a training sample set; processing each sample slice in the training sample set to generate aplurality of different artificial sample slices, and adding the training sample set after labeling the sample slices to expand the training sample set; constructing an SSD model; inputting the expanded training sample set into the constructed SSD model, and using the gradient descent method to train the SSD model; cutting the data to be measured into a plurality of slices which are the same sizeas the sample slices, and inputting the trained SSD model to obtain the target detection results of the data to be measured. The method of the invention has the advantages of strong robustness, fast running speed, high detection performance and easy migration, and can be used for detection under complex scenes without high contrast between the target and the background.

Description

Technical field [0001] The invention relates to the technical field of image processing, in particular to a method for detecting a SAR image iron tower target based on deep learning. Background technique [0002] Spaceborne Synthetic Aperture Radar (SAR) is a kind of microwave imaging radar. Its characteristic is that it is not affected by weather and climate. It can observe all-time, all-weather, high-resolution and large-area earth observation. Therefore, it is used in military target detection, marine Many fields such as monitoring, resource exploration, agriculture and forestry are widely used. [0003] Convolutional Neural Network (CNN) is a commonly used deep, pre-feedback artificial neural network. It is a kind of deep learning method and has been successfully applied in the field of computer vision. With the recovery of deep learning technology, target detection methods based on deep learning have developed rapidly in recent years. The R-CNN method based on the candidate ...

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): G06T7/00G06N3/04G06N3/08
CPCG06N3/08G06T7/0002G06T2207/20081G06T2207/10044G06N3/045
Inventor 李春升高原杨威陈杰
Owner BEIHANG 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