SAR image target recognition method based on cs and svm decision-level fusion
A decision-level fusion and target recognition technology, which is applied in the field of SAR image target recognition based on CS and SVM decision-level fusion, can solve problems such as the impact of classification accuracy, and achieve the effect of improving recognition rate and improving comprehensive ability
Active Publication Date: 2017-06-06
HANGZHOU DIANZI UNIV
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Decision-level fusion does not limit the selection of classifiers, but choosing an appropriate classifier has a certain impact on classification accuracy
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
[0041] The present invention will be further described below in conjunction with accompanying drawing.
[0042] Such as figure 1 Shown, the present invention comprises the following steps:
[0043] Step (1). Preprocessing
[0044] 1.1 Azimuth mark
[0045] Since the properties of the same target are very different under different azimuth angles, the directionality of the target should be considered in target recognition. Estimating the target azimuth can greatly reduce the number of features required for target recognition, thereby reducing the recognition time and improving the recognition performance. Manually mark the azimuth, and record the angle between the target and the standard azimuth (90 degrees). A mapping relationship is established between the marked azimuth and the index of the image, that is, a one-to-one correspondence between the azimuth and the index of the image, so that the azimuth of the image can be found through the index.
[0046] 1.2 Denoising and...
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 invention discloses an SAR (Synthetic Aperture Radar) image target recognition method based on CS (Compressed Sensing) and SVM (Support Vector Machine) decision fusion. According to the method, the respective advantages of CS and an SVM are combined; the optimization solving data of the CS is used for correcting an azimuth angle; and the recognition result of the CS and the SVM is subjected to decision fusion. Firstly, the SAR image target recognition problem is converted into the sparse signal recovery problem; the target classification result and the target azimuth angle estimation are respectively obtained on the basis of the recovered sparse coefficient; then, test images are subjected to posture correction, and the SVM is used for obtaining the target classification result; and finally, the three classification results are subjected to decision fusion according to a voting method. The experiment result shows that under the condition that the posture correction is not carried out, compared with other algorithms, a target recognition algorithm based on the CS has the advantages that the SAR image deformation target recognition accuracy is obviously improved; and under the condition that samples are few, the SAR deformation target recognition rate is obviously improved by the method provided by the invention.
Description
technical field [0001] The invention belongs to the technical field of target recognition, and relates to a SAR image target recognition method based on CS and SVM decision-level fusion. Background technique [0002] SAR (Synthetic Aperture Radar, Synthetic Aperture Radar) is a microwave imaging sensor that has a certain penetration ability to soil and vegetation, and has the characteristics of all-day, all-weather, multi-band, multi-polarization and high-resolution imaging. It has been widely used in economy and national defense construction. Research on Automatic Target Recognition (ATR) technology in SAR images, especially the recognition of deformed targets, is one of the key issues that need to be solved urgently. [0003] The process of SAR image target recognition can be described as: Find ROIS (Region of Interests, Region of Interest) from the image obtained by SAR observation, and then classify each ROIS to determine its category. The current research methods are ...
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 Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/66G06K9/00G06T5/00
Inventor 谷雨张琴彭冬亮陈华杰
Owner HANGZHOU DIANZI UNIV