In a wide-angle synthetic aperture radar (SAR), as the isotropic scattering assumption becomes invalid, scattering characteristics of targets tend to be aspect-dependent. In this paper, based on the high center frequency characteristic of Video SAR (ViSAR), a ViSAR aspect-dependent scattering imaging algorithm (VADSIA) is proposed, which includes two main procedures of persistent polar format algorithm (PPFA) and generalized likelihood ratio test (GLRT). Through PPFA, sequential ViSAR sub-aperture images (sub-images) can be formed to the same orientation with the same image resolution efficiently, providing the advantage of easy extraction and display of the aspect-dependent scattering characteristic. Different from conventional algorithms that require 2-D aspect-dependent interpolation procedures, PPFA utilizes a uniform lookup table scheme and a 2-D Chirp-Z transform (2D-CZT) to realize persistent sub-aperture imaging. As PPFA includes only 2-D fast Fourier transforms and multiplications, it is suitable for parallel implementation to realize real-time imaging. Besides, with the persistent imaging ability of PPFA, an efficient approximate GLRT process can be applied on the formed sub-images directly, which can preserve both the aspect-dependent scattering characteristics and the azimuth resolution of targets at the same time. Combining PPFA with GLRT, VADSIA is an effective and efficient way for the aspect-dependent imaging. The ground-based experimental results validate the effectiveness of the proposed algorithm.
Early recognition of gestures gives a better user experience during human–computer interaction in a near real-time way. However, the inadequate information accumulated in the early stages of the gesture will make the machine ambiguous and respond wrongly. Therefore, the challenge of early recognition lies in choosing the right time to trigger gesture prediction and using limited information to achieve reliable accuracy. In this article, an early gesture recognition method is proposed to achieve almost the same accuracy as a complete gesture sequence. For this purpose, we design a new architecture to modify the current feature sequence according to the existing information and introduce a new loss function to maximize the probability of correct gesture label as early as possible. At the same time, gesture recognition is triggered when the maximum probability of model output is greater than a set threshold. The publicly available data set used for evaluation comes from Google’s Project Soli sensor. Experimental results show that the proposed method for early recognition can achieve a gesture recognition rate of 88.85% with an average time saving of 69.47%. Compared with the results of all sequence recognition, the recognition rate drops by only 1.7%. In addition, the proposed method based on the data set constructed by terahertz radar also shows similar early recognition performance and can be applied to other radar sensors.
28 December 2018 in IEEE Geoscience and Remote Sensing Letters
Video synthetic aperture radar (ViSAR) is a novel imaging model to persistently surveil the region of interest by a series of consecutive images in real time. The formation time of one frame image, which can influence the image latency, is one of the key factors to realize ViSAR system. In general, the polar format algorithm (PFA) is preferred for ViSAR imaging. However, the 2-D resampling in the PFA has a large computation burden and is complex to realize in the real-time processing system. It has become the main obstacle to reduce image latency. Therefore, we derive four judgment criteria in this letter. According to these criteria, we propose the improved method of the PFA. Based on this method, the modified PFA can reduce the image formation time while guaranteeing the image quality. Simulation results are presented to validate the effectiveness of the improved method proposed in this letter.
11 September 2020 in IEEE Transactions on Instrumentation and Measurement
Range and Doppler frequency cells migration induced by the movement of low-observable target would bring about serious integration loss, which may deteriorate the target detection or imaging performance. Hence, this paper proposes a migration correction algorithm based on keystone transform and modified generalized Radon Fourier transform (KT-MGRFT). More specifically, KT is first utilized to eliminate the linear range cell migration (RCM) induced by the target unambiguous velocity. Then, MGRFT is presented to estimate the fold factor and acceleration. After that, the residual RCM and Doppler frequency cell migration (DFCM) are compensated by multiplying a reference signal constructed by the estimated fold factor and acceleration. Finally, the coherent integration can be achieved via inverse Fourier transform (IFT) and Fourier transform (FT) with respect to range frequency and scaled slow time, respectively. Compared with standard GRFT, the proposed method can remove the searching procedure of unambiguous velocity and thus it has lower computational complexity. Additionally, through taking the maximum value in searching range-Doppler frequency domain, the blind speed side-lobe (BSSL) effect is suppressed. Both the numerical simulations and the real experiment with Xband radar are given to verify the effectiveness of the proposed method.
29 June 2022 in IEEE Transactions on Instrumentation and Measurement
Terahertz (THz) radar system can realize high-resolution imaging benefits from its advantages, such as wide bandwidth and short synthetic aperture time. However, due to the short wavelength in the THz frequency range, multi-scattering center effect of targets generally occurs. This paper considers the mismatch between imaging aperture and target characteristics leads to serious resolution degradation, imaging defocus and loss of target details. We propose an aperture adaptive scale transform (AAST) imaging method, which is implemented in three steps. First, the sub-aperture is adaptively divided by matching the azimuth with the scattering characteristics. Then the apertures which contain different details of the target are scaled by weight. Finally, the generalized likelihood ratio test method is used to reconstruct the fine image results of incoherent scale transformation, resulting in more detailed information and higher definition in imaging results. We use a terahertz radar system operating at 0.34 THz and with a bandwidth of 28.8 GHz to scan around a target. The imaging results of measured echo corroborate the validity of the proposed algorithm.
30 September 2024 in IEEE Transactions on Geoscience and Remote Sensing
Terahertz (THz) circular synthetic aperture radar (CSAR) demands higher precision in motion compensation, posing significant challenges. To address this issue, this paper proposes a multi-channel motion compensation algorithm based on adaptive sub-aperture division (MCAAD). First, the Adaptive Aperture Division (AAD) algorithm is used to divide the image into multiple sub-apertures. Within each sub-aperture, initial coarse compensation is performed using data recorded by the inertial navigation system (GPS/INS). Subsequently, high-frequency vibration errors are compensated using the interferometric phase of different channels. Then, an image reconstruction algorithm is employed to image the compensated results, and a residual error estimation algorithm is applied to compensate for high-order phase errors and residual envelope errors (REE), yielding well-focused sub-aperture images. Finally, the sub-aperture images are fused using a base-4 registration algorithm. Simulation and experimental results with measured data demonstrate that the proposed algorithm effectively compensates for terahertz CSAR imaging.