Zheng Fang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Zheng Fang has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Gong Zhang
Nanjing University of Aeronautics and Astronautics
10 shared publications - Qijun Dai
Key Lab of Radar Imaging and Microwave Photonics, Ministry of Education, Nanjing University of Aeronautics and Astronautics, Nanjing, China
7 shared publications - Biao Xue
Xiamen University, Xiamen, China
6 shared publications - Peng Wang
Nanjing University of Aeronautics and Astronautics
3 shared publications Biao Xue
School of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
2 shared publications- Henry Leung
University of Calgary, Canada
2 shared publications
Latest Publications
- 4 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 24 (9) , 14493-14502
Power Line Detection Based on Subpixel-Neighborhood Attention in Complex Terrain Backgrounds
Helicopter crashes caused by collisions with power lines during low-altitude flights have been reported frequently. This paper proposes a new power line detection method based on Subpixel-Neighborhood Attention (SNA) to improve the detection performance of power lines in complex terrain backgrounds for helicopter flight safety. In complex terrain backgrounds, power line distribution tends to be sparse and power lines lack distinct features. This leads to intermittent detection results, loss of detailed information, and poor detection performance. Subpixel-Neighborhood Attention interpolates the neighborhood of Neighborhood Attention to refine it, obtaining high-resolution subpixel neighborhoods before conducting self-attention operations. As a result, SNA establishes connections between various pixels of power lines, producing continuous and more accurate detection results while improving the ability to extract detailed information about power lines in low-resolution images. The experiments on the Transmission Line Aerial-image dataset demonstrate that adding SNA to the power line detection method significantly improves the detection accuracy of power lines by 7.82%, effectively enhancing the detection performance of helicopters in identifying power lines in complex terrain backgrounds.
- 4 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 24 (9) , 14493-14502
Power Line Detection Based on Subpixel-Neighborhood Attention in Complex Terrain Backgrounds
Helicopter crashes caused by collisions with power lines during low-altitude flights have been reported frequently. This paper proposes a new power line detection method based on Subpixel-Neighborhood Attention (SNA) to improve the detection performance of power lines in complex terrain backgrounds for helicopter flight safety. In complex terrain backgrounds, power line distribution tends to be sparse and power lines lack distinct features. This leads to intermittent detection results, loss of detailed information, and poor detection performance. Subpixel-Neighborhood Attention interpolates the neighborhood of Neighborhood Attention to refine it, obtaining high-resolution subpixel neighborhoods before conducting self-attention operations. As a result, SNA establishes connections between various pixels of power lines, producing continuous and more accurate detection results while improving the ability to extract detailed information about power lines in low-resolution images. The experiments on the Transmission Line Aerial-image dataset demonstrate that adding SNA to the power line detection method significantly improves the detection accuracy of power lines by 7.82%, effectively enhancing the detection performance of helicopters in identifying power lines in complex terrain backgrounds.
- 10 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 20 (1545598X) , 1-5
Capsule-Guided Multi-View Attention Network for SAR Target Recognition With Small Training Set
Although numerous classification algorithms based on deep learning can effectively extract valuable features from synthetic aperture radar (SAR) images to improve the accuracy of SAR automatic target recognition, most of them have rigorous constraints in reality and fail to obtain satisfying results with limited training samples. To address the above issues, we propose a novel multi-view attention capsule network for SAR target small sample recognition. In our method, after gaining the primary capsules of SAR images under different views, a capsule-based view attention module is designed to enhance the feature relations between different views. Then a joint dynamic routing mechanism is adopted to further capture the robust inter- and intra-image spatial relations to generate SAR capsules specialized in the final classification. Especially, to alleviate the impact of SAR target angle sensitivity on recognition, rotated cropping is applied to the original SAR images in advance. Finally, experimental results on the moving and stationary target recognition (MSTAR) and OpenSARShip dataset have demonstrated the superiority and robustness of the proposed method upon limited labeled samples.
- 10 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 20 (1545598X) , 1-5
Capsule-Guided Multi-View Attention Network for SAR Target Recognition With Small Training Set
Although numerous classification algorithms based on deep learning can effectively extract valuable features from synthetic aperture radar (SAR) images to improve the accuracy of SAR automatic target recognition, most of them have rigorous constraints in reality and fail to obtain satisfying results with limited training samples. To address the above issues, we propose a novel multi-view attention capsule network for SAR target small sample recognition. In our method, after gaining the primary capsules of SAR images under different views, a capsule-based view attention module is designed to enhance the feature relations between different views. Then a joint dynamic routing mechanism is adopted to further capture the robust inter- and intra-image spatial relations to generate SAR capsules specialized in the final classification. Especially, to alleviate the impact of SAR target angle sensitivity on recognition, rotated cropping is applied to the original SAR images in advance. Finally, experimental results on the moving and stationary target recognition (MSTAR) and OpenSARShip dataset have demonstrated the superiority and robustness of the proposed method upon limited labeled samples.
- 1 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 61 (01962892) , 1-16
Waveform Diversity Design of OFDM Chirp for Miniature Millimeter-Wave MIMO Radar Based on Dechirp
The orthogonal waveform diversity design and efficient hardware implementation are important issues in miniature multiple-input multiple-output (MIMO) radars. The orthogonal frequency division multiplexing (OFDM) chirp waveform has received attention recently because of its large time-bandwidth product, constant modulus, no range-Doppler coupling, good orthogonality, and good Doppler tolerance. The dechirp-on-receive technique can reduce the amount of raw sampled data in near-field miniature millimeter-wave (mmW) MIMO radar detection and synthetic aperture radar (SAR) imaging. However, most of the current waveform diversity design methods are based on general matched filtering (MF). In this paper, the possibility of using the traditional OFDM chirp waveform for dechirp processing at the receiving end of MIMO radar is analyzed. Then, the results of different configurations of chirp rates within and between transmitted waveforms for different signal processing procedures are investigated. A novel dechirp-based OFDM chirp waveform diversity design method for MIMO radar is proposed, and the results of the waveform design are given. Numerical results, such as pulse compression (PC) results, dechirp ambiguity function (DAF), SAR imaging processing, etc., and experiments verify the effectiveness of the proposed methods.
- 1 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 61 (01962892) , 1-16
Waveform Diversity Design of OFDM Chirp for Miniature Millimeter-Wave MIMO Radar Based on Dechirp
The orthogonal waveform diversity design and efficient hardware implementation are important issues in miniature multiple-input multiple-output (MIMO) radars. The orthogonal frequency division multiplexing (OFDM) chirp waveform has received attention recently because of its large time-bandwidth product, constant modulus, no range-Doppler coupling, good orthogonality, and good Doppler tolerance. The dechirp-on-receive technique can reduce the amount of raw sampled data in near-field miniature millimeter-wave (mmW) MIMO radar detection and synthetic aperture radar (SAR) imaging. However, most of the current waveform diversity design methods are based on general matched filtering (MF). In this paper, the possibility of using the traditional OFDM chirp waveform for dechirp processing at the receiving end of MIMO radar is analyzed. Then, the results of different configurations of chirp rates within and between transmitted waveforms for different signal processing procedures are investigated. A novel dechirp-based OFDM chirp waveform diversity design method for MIMO radar is proposed, and the results of the waveform design are given. Numerical results, such as pulse compression (PC) results, dechirp ambiguity function (DAF), SAR imaging processing, etc., and experiments verify the effectiveness of the proposed methods.
- 3 June 2023
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 14 (6) , 631-640
CVGAN: A Cross-View SAR Image Generation Method for Enhancing the View Diversity
- 3 June 2023
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 14 (6) , 631-640
CVGAN: A Cross-View SAR Image Generation Method for Enhancing the View Diversity
- Open Access
- 16 January 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (2) , 526
Hybrid Attention-Based Encoder–Decoder Fully Convolutional Network for PolSAR Image Classification
Recently, methods based on convolutional neural networks (CNNs) achieve superior performance in polarimetric synthetic aperture radar (PolSAR) image classification. However, the current CNN-based classifiers follow patch-based frameworks, which need input images to be divided into overlapping patches. Consequently, these classification approaches have the drawback of requiring repeated calculations and only relying on local information. In addition, the receptive field size in conventional CNN-based methods is fixed, which limits the potential to extract features. In this paper, a hybrid attention-based encoder–decoder fully convolutional network (HA-EDNet) is presented for PolSAR classification. Unlike traditional CNN-based approaches, the encoder–decoder fully convolutional network (EDNet) can use an arbitrary-size image as input without dividing. Then, the output is the whole image classification result. Meanwhile, the self-attention module is used to establish global spatial dependence and extract context characteristics, which can improve the performance of classification. Moreover, an attention-based selective kernel module (SK module) is included in the network. In the module, softmax attention is employed to fuse several branches with different receptive field sizes. Consequently, the module can capture features with different scales and further boost classification accuracy. The experiment results demonstrate that the HA-EDNet achieves superior performance compared to CNN-based and traditional fully convolutional network methods.
- 16 January 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (2) , 526
Open AccessHybrid Attention-Based Encoder–Decoder Fully Convolutional Network for PolSAR Image Classification
Recently, methods based on convolutional neural networks (CNNs) achieve superior performance in polarimetric synthetic aperture radar (PolSAR) image classification. However, the current CNN-based classifiers follow patch-based frameworks, which need input images to be divided into overlapping patches. Consequently, these classification approaches have the drawback of requiring repeated calculations and only relying on local information. In addition, the receptive field size in conventional CNN-based methods is fixed, which limits the potential to extract features. In this paper, a hybrid attention-based encoder–decoder fully convolutional network (HA-EDNet) is presented for PolSAR classification. Unlike traditional CNN-based approaches, the encoder–decoder fully convolutional network (EDNet) can use an arbitrary-size image as input without dividing. Then, the output is the whole image classification result. Meanwhile, the self-attention module is used to establish global spatial dependence and extract context characteristics, which can improve the performance of classification. Moreover, an attention-based selective kernel module (SK module) is included in the network. In the module, softmax attention is employed to fuse several branches with different receptive field sizes. Consequently, the module can capture features with different scales and further boost classification accuracy. The experiment results demonstrate that the HA-EDNet achieves superior performance compared to CNN-based and traditional fully convolutional network methods.
- 13 February 2022
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 13 (4) , 383-393
A hybrid network for PolSAR image classification based on polarization orientation angles
Polarimetric synthetic aperture radar (PolSAR) image classification is a significant task of PolSAR applications. In recent years, deep learning-based methods using spatial features have achieved satisfactory results in PolSAR classification. However, there are still some important features left to be exploited. One is the target scattering orientation diversity feature that contains rich hidden information. In this letter, a hybrid network is proposed to model the target scattering orientation diversity feature and spatial feature. First, the polarimetric scoherency matrix is expanded to polarization coherency matrix sequence by different polarization orientation angles (POAs). Then, the long short-term memory (LSTM) network and the convolutional neural networks (CNN) are proposed to process the polarization coherency matrix sequence and the square neighbourhood of pixels, respectively. Finally, the features of the two sub-networks are combined to improve the classification accuracy. The experiments demonstrate that the proposed method can obtain superior and robust performance.
- 13 February 2022
- Published by Taylor & Francis in Remote Sensing Letters
- Vol. 13 (4) , 383-393
A hybrid network for PolSAR image classification based on polarization orientation angles
Polarimetric synthetic aperture radar (PolSAR) image classification is a significant task of PolSAR applications. In recent years, deep learning-based methods using spatial features have achieved satisfactory results in PolSAR classification. However, there are still some important features left to be exploited. One is the target scattering orientation diversity feature that contains rich hidden information. In this letter, a hybrid network is proposed to model the target scattering orientation diversity feature and spatial feature. First, the polarimetric scoherency matrix is expanded to polarization coherency matrix sequence by different polarization orientation angles (POAs). Then, the long short-term memory (LSTM) network and the convolutional neural networks (CNN) are proposed to process the polarization coherency matrix sequence and the square neighbourhood of pixels, respectively. Finally, the features of the two sub-networks are combined to improve the classification accuracy. The experiments demonstrate that the proposed method can obtain superior and robust performance.
- 27 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
PolSAR Image Classification Based on Complex-Valued Convolutional Long Short-Term Memory Network
Polarimetric synthetic aperture radar (PolSAR) image classification is an essential part of PolSAR image interpretation. In recent years, convolutional neural networks (CNN) have made significant advances in PolSAR image classification. However, the current CNN-based methods ignore complementary information among different feature maps and correlations between elements of coherence matrix, which can provide discriminative information for classification. Besides, the phase information contained in the complex-valued (CV) coherence matrix cannot be extracted effectively. In this letter, a stacked CV convolutional long short-term memory network (ConvLSTM) called CV-ConvLSTM is proposed for PolSAR classification. Compared to existing methods, the CV-ConvLSTM can extract complementary information among different feature maps and utilize the dependencies of elements in the coherency matrix, which can improve the performance of classification. In addition, the CV operations are added to the network, in which phase information is used for better classification. The experiments results of two widely used PolSAR datasets demonstrate that the CV-ConvLSTM can obtain superior performance compared with existing CNN methods.
- 27 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
PolSAR Image Classification Based on Complex-Valued Convolutional Long Short-Term Memory Network
Polarimetric synthetic aperture radar (PolSAR) image classification is an essential part of PolSAR image interpretation. In recent years, convolutional neural networks (CNN) have made significant advances in PolSAR image classification. However, the current CNN-based methods ignore complementary information among different feature maps and correlations between elements of coherence matrix, which can provide discriminative information for classification. Besides, the phase information contained in the complex-valued (CV) coherence matrix cannot be extracted effectively. In this letter, a stacked CV convolutional long short-term memory network (ConvLSTM) called CV-ConvLSTM is proposed for PolSAR classification. Compared to existing methods, the CV-ConvLSTM can extract complementary information among different feature maps and utilize the dependencies of elements in the coherency matrix, which can improve the performance of classification. In addition, the CV operations are added to the network, in which phase information is used for better classification. The experiments results of two widely used PolSAR datasets demonstrate that the CV-ConvLSTM can obtain superior performance compared with existing CNN methods.
- 10 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
An Applied Ambiguity Function Based on Dechirp for MIMO Radar Signal Analysis
The orthogonal waveform design and good hardware realization of multiple-input–multiple-output (MIMO) radar have always been important research topics. The orthogonal frequency division multiplexing (OFDM) chirp waveform has received more attention because of its large time-bandwidth product, constant modulus, no range-Doppler coupling, good orthogonality, and good Doppler tolerance. Dechirp technique can reduce the amount of raw sampled data very well in near-field miniature lightweight MIMO radar detection and synthetic aperture radar (SAR) imaging. However, most of the current waveform analysis methods are based on matched filtering (MF). In this letter, an ambiguity function (AF) based on the dechirp signal processing approach to analyze the waveform performance is proposed, called dechirp ambiguity function (DAF). The pulse compression performance of the waveform itself and the level of mutual interference between the waveforms are described from the perspective of DAF. Numerical results validate reliability and effectiveness of the DAF.
- 10 January 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
An Applied Ambiguity Function Based on Dechirp for MIMO Radar Signal Analysis
The orthogonal waveform design and good hardware realization of multiple-input–multiple-output (MIMO) radar have always been important research topics. The orthogonal frequency division multiplexing (OFDM) chirp waveform has received more attention because of its large time-bandwidth product, constant modulus, no range-Doppler coupling, good orthogonality, and good Doppler tolerance. Dechirp technique can reduce the amount of raw sampled data very well in near-field miniature lightweight MIMO radar detection and synthetic aperture radar (SAR) imaging. However, most of the current waveform analysis methods are based on matched filtering (MF). In this letter, an ambiguity function (AF) based on the dechirp signal processing approach to analyze the waveform performance is proposed, called dechirp ambiguity function (DAF). The pulse compression performance of the waveform itself and the level of mutual interference between the waveforms are described from the perspective of DAF. Numerical results validate reliability and effectiveness of the DAF.
- 1 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
SAR Target Recognition With Modified Convolutional Random Vector Functional Link Network
Deep learning models have achieved remarkable performance in synthetic aperture radar (SAR) target recognition. However, the accuracy of these methods is sensitive to the hyper-parameters and the traditional backpropagation is time consuming. In this letter, we proposed a modified convolutional random vector functional link (IntCRVFL) network for SAR target recognition, which can simplify the SAR target recognition system. The convolutional random vector functional link network consists of a convolutional neural network and a random vector functional link (RVFL) network. First, the fixed convolutional layers with randomly initialized parameters extract SAR image features and then the RVFL network performs target recognition. Especially, inspired by hyperdimensional computing, the activations of the hidden layer are obtained through a new encoding manner. Besides, only the connections between hidden and output layers need to train by a closed-form solution for the ultimately precise target recognition. The experimental results on the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set demonstrate the proposed IntCRVFL network can obtain a satisfying accuracy with a faster speed.
- 1 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
SAR Target Recognition With Modified Convolutional Random Vector Functional Link Network
Deep learning models have achieved remarkable performance in synthetic aperture radar (SAR) target recognition. However, the accuracy of these methods is sensitive to the hyper-parameters and the traditional backpropagation is time consuming. In this letter, we proposed a modified convolutional random vector functional link (IntCRVFL) network for SAR target recognition, which can simplify the SAR target recognition system. The convolutional random vector functional link network consists of a convolutional neural network and a random vector functional link (RVFL) network. First, the fixed convolutional layers with randomly initialized parameters extract SAR image features and then the RVFL network performs target recognition. Especially, inspired by hyperdimensional computing, the activations of the hidden layer are obtained through a new encoding manner. Besides, only the connections between hidden and output layers need to train by a closed-form solution for the ultimately precise target recognition. The experimental results on the Moving and Stationary Target Acquisition and Recognition (MSTAR) data set demonstrate the proposed IntCRVFL network can obtain a satisfying accuracy with a faster speed.
- 17 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
Semisupervised Deep Convolutional Neural Networks Using Pseudo Labels for PolSAR Image Classification
Deep-learning-based methods have obtained satisfying results in polarimetric synthetic aperture radar (PolSAR) image classification. However, these methods require large numbers of labeled samples, which are usually time-consuming and high-priced for PolSAR images. To address this issue, a semisupervised method based on a 3-D convolutional neural network (3-D-CNN) using pseudo labels (PL-3-D-CNN) is proposed. First, the coherency matrix of PolSAR data is converted into a 6-D real-valued vector by a unitary transformation. Then, the K-means algorithm is utilized for generating pseudo labels. After that, labeled samples and pseudo labeled samples are fed into the PL-3-D-CNN model to extract supervised and unsupervised features. Finally, the supervised and unsupervised features are combined to improve classification accuracy. The proposed method is tested on both AIRSAR and RADARSAT-2 data sets. The results show that the proposed method is an effective method for PolSAR image classification and shows good performance under a small number of labeled samples. The source code for the PL-3-D-CNN model is available at https://github.com/fangzheng-nuaa/PL-3D-CNN.
- 17 November 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 19 (1545598X) , 1-5
Semisupervised Deep Convolutional Neural Networks Using Pseudo Labels for PolSAR Image Classification
Deep-learning-based methods have obtained satisfying results in polarimetric synthetic aperture radar (PolSAR) image classification. However, these methods require large numbers of labeled samples, which are usually time-consuming and high-priced for PolSAR images. To address this issue, a semisupervised method based on a 3-D convolutional neural network (3-D-CNN) using pseudo labels (PL-3-D-CNN) is proposed. First, the coherency matrix of PolSAR data is converted into a 6-D real-valued vector by a unitary transformation. Then, the K-means algorithm is utilized for generating pseudo labels. After that, labeled samples and pseudo labeled samples are fed into the PL-3-D-CNN model to extract supervised and unsupervised features. Finally, the supervised and unsupervised features are combined to improve classification accuracy. The proposed method is tested on both AIRSAR and RADARSAT-2 data sets. The results show that the proposed method is an effective method for PolSAR image classification and shows good performance under a small number of labeled samples. The source code for the PL-3-D-CNN model is available at https://github.com/fangzheng-nuaa/PL-3D-CNN.