Based on momentum method BP neural network in the target recognition research and application

Xue-feng Zhang; Yu-bin Gao

doi:10.1117/12.900494

8 September 2011 Based on momentum method BP neural network in the target recognition research and application

Xue-feng Zhang, Yu-bin Gao

Proceedings Volume 8193, International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Infrared Imaging and Applications; 81932N (2011) https://doi.org/10.1117/12.900494
Event: International Symposium on Photoelectronic Detection and Imaging 2011, 2011, Beijing, China

Abstract

Target recognition is measured by treating target existing knowledge to judge, analyze, and thus the process of target identification. Using anti-vibration lateral shearing interferometer to get the interference fringe for the spectrum information of measurement target, and the system can get the target by spectrum identification algorithm. By the condition that interferometer's length isn't changed, the system was optimized by momentum BP Neural Network algorithm in the separating mixed spectrum process, therefore it could improve the probability of camouflage target recognition. The spectrum information was calculated by the fringes, to getting the mixed spectrum data. The absorption spectrum was in the hidden layer, and the system obtained every kinds of characteristic spectrum from mixed spectrum by the momentum BP Neural Network. Experiments showed that it collected mixed spectrum of background form different distances and different surface, and made them to the initial spectrum information. The test target was a board that it's surface was made to four kinds, and there was no paint (A), brushing camouflage paint of military green (B), brushing camouflage paint of irregular shape (C) and brushing camouflage paint of irregular box (D). The mixed spectrum was obtained from the anti-vibration lateral shearing interferometer, while the recognition probability for non-camouflage target were above 90.0% by the traditional algorithm and the momentum BP neural network algorithm, but the recognition probability for camouflage target was 85.6% by momentum BP neural network algorithm, better than 41.5% by the traditional algorithm, so it proved that the algorithm could improve the recognition probability for camouflage target effectively.

Citation Download Citation

Xue-feng Zhang and Yu-bin Gao "Based on momentum method BP neural network in the target recognition research and application", Proc. SPIE 8193, International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Infrared Imaging and Applications, 81932N (8 September 2011); https://doi.org/10.1117/12.900494

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available