Target Recognition in SAR Images Using Complex-Valued Network Guided with Sub-Aperture Decomposition
نویسندگان
چکیده
Synthetic aperture radar (SAR) images have special physical scattering characteristics owing to their unique imaging mechanism. Traditional deep learning algorithms usually extract features from real-valued SAR in a purely data-driven manner, which may ignore some important and sacrifice useful target information images. This undoubtedly limits the improvement performance for recognition. To take full advantage of contained images, complex-valued network guided with sub-aperture decomposition (CGS-Net) recognition is proposed. According fact that different targets at angles, used improve accuracy multi-task strategy. Specifically, proposed method includes main auxiliary tasks, can task by sharing task. Here, task, reconstruction In addition, original effectively utilizes amplitude phase The experimental results obtained using MSTAR dataset illustrate CGS-Net achieved an 99.59% (without transfer or data augmentation) ten-classes targets, superior other popular methods. Moreover, has lightweight structure, suitable tasks because lack large number labeled data. small further demonstrate excellent CGS-Net.
منابع مشابه
Target Detection In Sar Images Based On Sub-Aperture Coherence And Phase Congruency
For target detection in SAR images, the sub-aperture coherence analysis is employed widely by calculating coefficient of coherence to express the differences of the target signals in sub-aperture images. However the calculation of coherence coefficients is non-adaptive so that when the amplitude difference of coherence coefficient between a target and background is small target detection probab...
متن کاملMoving-target Velocity Estimation in a Complex-valued Sar Imagery
In the ground moving target indication with synthetic aperture radar (SAR) community, algorithms used to estimate the velocity of a detected moving target are important because they are relative to the topics about refocusing and azimuth displacement correction. The velocity is regarded as a vector with two components, one in azimuth and one in range direction, and new algorithms aiming at esti...
متن کاملTarget Recognition for Multi-aspect Sar Images with Fusion Strategies
Two fusion strategies for target recognition using multiaspect synthetic aperture radar (SAR) images are presented for recognizing ground vehicles in MSTAR database. Due to radar crosssection variability, the ability to discriminate between targets varies greatly with target aspect. Multi-aspect images of a given target are used to support recognition. In this paper, two fusion strategies for t...
متن کاملSAR Target Recognition Using Improved Fuzzy Neural Network
Target recognition in high-resolution synthetic aperture radar (SAR) images is a challenging task, because SAR images have higher ambiguity for different target, which will reduce the correct recognition rate. This paper presents an improved SAR recognition algorithm based on fuzzy neural network (FNN), which deals with the ambiguity SAR target recognition very well. This improved FNN system im...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Remote Sensing
سال: 2023
ISSN: ['2315-4632', '2315-4675']
DOI: https://doi.org/10.3390/rs15164031