Search

Funding Information: This study was supported by the National Natural Science Foundation of China (Grant No. 62001229, 62101264, 62101260) and by China Postdoctoral Science Foundation (Grant No. 2020M681604). O.A. was supported by Multico project funded by Business Finland and Forest Carbon Monitoring project funded by European Space Agency. Publisher Copyright: © 2022 by the authors. Time series of SAR imagery combined with reference ground data can be suitable for producing forest inventories. Copernicus Sentinel-1 imagery is particularly interesting for forest mapping because of its free availability to data users; however, temporal dependencies within SAR time series that can potentially improve mapping accuracy are rarely explored. In this study, we introduce a novel semi-supervised Long Short-Term Memory (LSTM) model, CrsHelix-LSTM, and demonstrate its utility for predicting forest tree height using time series of Sentinel-1 images. The model brings three important modifications to the conventional LSTM model. Firstly, it uses a Helix-Elapse (HE) projection to capture the relationship between forest temporal patterns and Sentinel-1 time series, when time intervals between datatakes are irregular. A skip-link based LSTM block is introduced and a novel backbone network, Helix-LSTM, is proposed to retrieve temporal features at different receptive scales. Finally, a novel semisupervised strategy, Cross-Pseudo Regression, is employed to achieve better model performance when reference training data are limited. CrsHelix-LSTM model is demonstrated over a representative boreal forest site located in Central Finland. A time series of 96 Sentinel-1 images are used in the study. The developed model is compared with basic LSTM model, attention-based bidirectional LSTM and several other established regression approaches used in forest variable mapping, demonstrating consistent improvement of forest height prediction accuracy. At best, the achieved accuracy of forest height mapping was 28.3% relative root mean squared error (rRMSE) for pixel-level predictions and 18.0% rRMSE on stand level. We expect that the developed model can also be used for modeling relationships between other forest variables and satellite image time series.

The performance of synthetic aperture radar is vulnerable to radio frequency interference (RFI). In many situations, the RFI has a low-rank property, since the frequency bands occupied by RFI usually remain stable during a short slow time period. Therefore, low-rank representation (LRR)-based methods can be applied to separate RFI and signal of interest (SOI), by minimizing the rank of RFI components with a regularization constraint to protect SOI. However, traditional methods use the sparsity of the raw data or range profile to formulate the regularization term, which fails to describe the properties of SOI accurately. In addition to the sparse property of range profiles, this article explores the common patterns hidden in the range profiles and proposes two new LRR-based RFI suppression optimization models with a well-designed regularization term to describe such common sparsity to protect the SOI. Four methods are proposed to solve the optimization problems based on the alternating direction multiplier (ADM) method, which provides tradeoff between efficiency and accuracy. Compared with traditional LRR-based RFI suppression methods, the proposed methods make a more precise description of the features of SOI, therefore can better protect the information of SOI during the RFI suppression process and improves the imaging quality. The superior performance of the proposed method is validated by measured data in both sparse and nonsparse scenes.

Radio frequency interference (RFI) can severely degrade the image quality for synthetic aperture radar (SAR). Traditional sparse recovery-based RFI suppression methods construct a joint dictionary to represent both the time-domain RFI and the useful target echo (UTE), thus, transforming the task of RFI suppression into a signal recovery problem. However, these methods generally have poor performance when the one-dimensional (1-D) range profiles are nonsparse. To overcome this problem, two RFI suppression algorithms are proposed based on a modified block sparse Bayesian learning (BSBL). In the first algorithm, the vector to be recovered is divided into several blocks of identical size. The Bayesian hyper-parameters corresponding to the RFI and UTE are learned separately by exploiting the temporal intrablock correlation, and then the nonsparse vector can be successfully recovered. In the second algorithm, the dictionary is adaptively tuned during the iteration process of BSBL, thus reducing the computational load. After recovering the UTE, a well-focused 2-D image can be obtained by traditional SAR imaging algorithms. With the use of properties of both the RFI and UTE, and by exploiting the intrablock correlation, the proposed RFI suppression methods outperform other sparse recovery methods, especially when the range profile is nonsparse. Simulation results demonstrate the superior performance of the proposed algorithms.

Super-resolution technology is considered as an efficient approach to promote the image quality of forward-looking imaging radar. However, super-resolution technology is inherently an ill-conditioned issue, whose solution is quite susceptible to noise. Bayesian method can efficiently alleviate this issue through utilizing prior knowledge of the imaging process, in which the scene prior information plays a pretty significant role in ensuring the imaging accuracy. In this paper, we proposed a novel Bayesian super-resolution method on the basis of Markov random field (MRF) model. Compared with the traditional super-resolution method which is focused on one-dimensional (1-D) echo processing, the MRF model adopted in this study strives to exploit the two-dimensional (2-D) prior information of the scene. By using the MRF model, the 2-D spatial structural characteristics of the imaging scene can be well described and utilized by the nth-order neighborhood system. Then, the imaging objective function can be constructed through the maximum a posterior (MAP) framework. Finally, an accelerated iterative threshold/shrinkage method is utilized to cope with the objective function. Validation experiments using both synthetic echo and measured data are designed, and results demonstrate that the new MAP-MRF method exceeds other benchmarking approaches in terms of artifacts suppression and contour recovery.

With a long synthetic aperture time, the energy of the ground moving target will be defocused due to the effect of range migration (RM) and Doppler frequency migration (DFM). This paper combines Radon transform and the modified Lv's distribution (MLVD), then proposes RMLVD to eliminate RM and estimate the motion parameters of target. It is validated that the proposed method can estimate the equivalent slant acceleration and jerk simultaneously. Besides, the proposed method can realize coherent integration with a low computational load and eliminate the Doppler ambiguity effect. The simulation results show that the proposed method has an excellent detection performance in the multiple targets or low signal-to-noise ratio (SNR) environment.

Multiple-input multiple-output (MIMO) synthetic aperture radar (SAR) waveform diversity design has drawn extensive attention since it has a great potential in remote sensing. Waveform designing becomes one of the most significant tasks in the MIMO SAR system. In this study, a novel scheme has been proposed to generate four waveforms using an orthogonal frequency division multiplexing (OFDM) chirp waveform. The four waveforms have good auto-correlation properties, ambiguity performance, and a good peak sidelobe ratio. This scheme makes full use of the bandwidth. The orthogonal waveform is demodulated by padding zeros and under sampling in the frequency domain. Simulation results are obtained to validate that more orthogonal waveforms can be produced by our proposal scheme with a comparative performance.

This paper investigates the problem of simultaneous target detection and tracking with heavy-tailed noises under Bernoulli filtering framework. By exploiting the Student-t distributions to formulate the outlier-corrupted process noise and measurement noise, a novel target tracking algorithm named the Student-t mixture Bernoulli filter is proposed. Under linear assumptions on target dynamic model, sensor measurement model and birth process, closed-form recursions that propagate the parameters of Bernoulli random finite set are derived. The validity of the proposed algorithm is verified by numerical simulations.

Development of random signal radar (RSR) over the past 30 years is described. Conventional methods of implementing RSR are summarized such as correlation, spectrum analysis, and anticorrelation. Some typical RSR systems are introduced, for example, noise frequency modulation CW radar, random binary phase-coded CW radar, etc., and their merits and demerits are also pointed out. Finally, RSR development trends are analyzed.