Search

Track-before-detect (TBD) algorithm has been used to track weak moving target shadows in video synthetic aperture radar (SAR), where strong maneuverability may deteriorate tracking performance. Fortunately, the Doppler characteristic of the target can additionally provide velocity guidance for tracking. This article presents ground moving target indication (GMTI) results of the multichannel video SAR raw data, and a tracking approach is proposed based on the Doppler supervision in an improved dynamic programming-based TBD (DP-TBD) framework that uses a dual-domain merit function. Both the sequential high-resolution SAR images and low-resolution range-Doppler (RD) spectra are used, and the clutter in the RD spectrum is suppressed by using an adaptive displaced phase center antenna (ADPCA) technique. Fine states can be searched in dual-domain through resampling from random expansion in state initialization and transition. The inverse shadow amplitude and Doppler energy of the same potential target are simultaneously integrated to determine whether the target exists. The proposed Dual-DP-TBD can deal with the tracking of a time-varying number of targets in successive measurements. Compared to other TBD algorithms, this approach has fewer missing alarms and false alarms, thanks to precise velocity estimation constrained from diverse features both in SAR image and RD spectrum. [ABSTRACT FROM AUTHOR]

The spatially and temporally correlated sea clutter with phase information is valuable for marine radar applications. The major difficulty of coherent sea clutter modeling is the generation of the continuous phases. This article presents a new phase retrieval approach for modeling the correlated complex sea clutter based on unsupervised neural networks. The unsupervised short-term and long-term neural networks have been developed for the phase retrieval on different term scales. Both these networks have the same input layer and feature extraction module, and however, the number of output neurons is different. The amplitude sea clutter series and the desired Doppler spectrum are fed into the network in parallel, and their features are extracted by two parallel bidirectional long short-term memory (Bi-LSTM) networks which sufficiently utilize the correlations of sea clutter data. These features are concatenated and fused by a residual network (ResNet). The phases can be successfully obtained by constraining to the desired Doppler spectrum and the given amplitudes of sea clutter series. This proposed approach has been verified by the measured Ice Multiparameter Imaging X-Band (IPIX) radar data, and it can precisely model the complex sea clutter with specified statistic characteristics and Doppler properties. The amplitude root mean square error (RMSE) between the obtained and measured Doppler spectra is only 1.5065 with the interval between adjacent frames equals to 32. The RMSE of Doppler central frequency and spectrum width is 6.9306 and 1.2293 Hz, respectively. It shows robustness with the change of range resolution and interval. [ABSTRACT FROM AUTHOR]

It is challenging to recognize individuals when they move in the radar field of view due to the superimposition of micro-Doppler signatures. This paper presents a multi-person recognition approach by separating micro-Doppler signatures of multiple persons into their individual components. The preliminary separation can be obtained by their range difference in a high resolution radar. A multi-task learning network is designed for both the fine separation of micro-Doppler signatures and the personnel recognition. A frequency modulated continuous waveform (FMCW) radar that operates at 77 GHz for automotive applications is used in experiments. The proposed deep-neural-network-based approach gives a convincing result in the test. [ABSTRACT FROM AUTHOR]