Search

This paper presents vehicle occupant body imaging algorithms for detecting the presence of children that are left behind in vehicles. The proposed algorithms utilize occupancy grid mapping (OGM) techniques developed for autonomous driving, which are modified for in-cabin human detection purposes. Two different kinds of OGM techniques are presented to deal with two major difficulties in imaging the occupant bodies in an in-cabin environment. The processing flows of the proposed algorithms are detailed in this report. Two experimental results for 60 GHz millimeter-wave radars are also presented for performance verification.

A space-borne synthetic aperture radar (SAR) is widely used for maritime surveillance because it can observe a broad area. Although a low pulse repetition frequency (PRF) mode is generally selected to observe a wider area, it causes azimuth ambiguities that may lead to false detection and affect the accuracy of ship detection. In previous work, an azimuth ambiguity discrimination algorithm using the Doppler spectrum was presented. However, this algorithm did not eliminate false detection and presented challenges for ship detection performance. To achieve more precise detection, a novel algorithm using a compressive sensing (CS)-based SAR image is proposed. CS can reconstruct the ambiguity-suppressed SAR image, resulting in further improvement of azimuth ambiguity discrimination. The proposed algorithm was applied to a SAR image simulated by half-downsampling of the received signals. The evaluation shows that the proposed algorithm suppresses the power of azimuth ambiguity up to 15 dB and eliminates false detection. These findings demonstrate that the algorithm discriminates robustly and significantly improves ship detection performance.

SAR (Synthetic Aperture Radar) moving target along-track velocity estimation method for multichannel system is proposed. The proposed method works on the co-registered multichannel SAR images and employ sub-aperture processing to enhance the robustness to the clutter and noise. The velocity estimation performance of the proposed algorithm is shown through airborne Ku-band three channel SAR experiments, and it has been shown that the accuracy of the along track velocity estimation of moving target at the range of about 2.5km is on the order of 0.2 m/s.

An image formation processing of airborne synthetic aperture radar (SAR) assumes a focal plane for motion compensation (MoCo). When a focal plane is set on the ground, top of a high building might be blurred in the resulting SAR image. Therefore, we propose a focus stacking algorithm to make a focused airborne SAR image. This algorithm automatically combines some subdivided SAR images which have different focal planes each other. Autocorrelation is calculated at each pixel as azimuth focusing reference index to find the most focused image. It needs no external information such as digital surface model (DSM).

Range ambiguity level is one of the most important metrics of the spaceborne synthetic aperture radar (SAR) image quality. Alternating transmission of up- and down-chirp signals is an effective method to suppress the range ambiguity. However, this modulation method brings new unwanted azimuth ambiguity, since an intra-pulse Doppler effect causes opposite range shift for up- and down- chirp. In this paper, we propose a signal processing strategy for multi-channel sliding spotlight SAR with alternate up- and down-chirp modulation. The proposed algorithm is based on the sub-aperture method with azimuth digital beamforming, and a special care is taken for the correction of the intra-pulse Doppler effect to suppress the azimuth ambiguity. The performance of the intra-pulse Doppler effect correction process is evaluated through a point target simulation, and it is shown that the ambiguity-to-signal ratio (AASR) has been improved by 4.4dB.

We propose a novel SAR azimuth ambiguity suppression method. The proposed method is a post process for a SAR image; namely, it does not require any modification to the SAR image formation process nor the SAR system design. Noting that the azimuth ambiguity is blurred in the SAR image, the proposed algorithm refocuses the image to focus the azimuth ambiguity and then automatically detects and identifies the locations, and also the orders, of the ambiguities in the refocused image. After the detected ambiguities in the refocused image are eliminated, the final SAR image is obtained by refocusing back to the true image. Experimental results using true satellite images show that the proposed method successfully suppresses the ambiguities without causing severe degradations to the true images.

Conventional synthetic aperture radar (SAR) systems suffer from the contradictory requirements of high resolution and wide swath observation. SAR systems with multiple receive channels in azimuth and/or elevation have been studied and shown to overcome this trade-off. When the system with multiple receive channels in elevation employs variable pulse repetition frequency (PRF) in order to eliminate the blind ranges, it requires azimuth resampling process at some point in the image formation. If the system also has multiple channels in azimuth, then there will be an additional difficulty in the resampling process, because the average PRF is set to be lower than the Doppler bandwidth of the signal received by each receive channel.In this paper, we propose a processing strategy for combining the variable PRF SAR observation with multiple channels both in elevation and azimuth. In the proposed method, multiple azimuth beams are formed at first for each pulse, and azimuth resampling is applied to the signals of each beam. In order to reduce the resampling error, overlapped beams are formed, and only a fraction in the vicinity of the center of the beams are used to reconstruct the full Doppler spectrum.

Compressive sensing (CS) is based on the random coding theorem. In this term, lowness of the coherence of forward/backward transformation matrix is important. Random sampling makes coherence low, then conventional algorithms are based on random sampling. A novel idea to get the low coherency of ambiguity is not random sampling but simple decimation of the SAR azimuth arrays. To make simple decimated array, the SAR system only needs to set low pulse repetition frequency (PRF). In this paper, we have demonstrated our method of compressive sensing for synthetic aperture radar (SAR) is effective even in the case of sub-Nyquist linearly decimated array.

Synthetic Aperture Radar (SAR) can take a fine radar image without relation to sunshine condition. Recently, video generated from SAR is discussed. The radar video offers the practical understanding of the target motion without the added implement. The radar video is often generated from raw data in spotlight mode. In contrast, we propose the radar video generated from SAR image. The proposed scheme provides a low complexity and the freedom of the observation mode. In the paper, we show the process of the proposed scheme and indicate the snapshot image of radar video made of our air-borne stripmap SAR data or Advanced Land Observing Satellite-2 data. The images show possibility of the radar video with stripmap mode or space-borne SAR.

For urban area monitoring, detection and classification of small targets with synthetic aperture radar imagery are one of the useful solutions. The conventional method has the problem that when a target on both master and slave images at the same position has the same radar cross section, it does not classify the target as either a slightly moved reparked vehicle or as a different vehicle. To overcome this problem, we propose a Fourier-transform-based simple phase compensation method for both accurate coherence estimation and classification of the reparked vehicle's status. The proposed method obtains the maximum value of the frequency basis, and also determines the threshold, which is based on effective sample number, with a constant false alarm rate. This number is determined by the peak signal-to-noise ratio in estimation window, and are used on the preprocessed probability density function for each sample number. Finally, the proposed method classifies a target as either a slightly moved vehicle or as a different vehicle by using both the proposed coherence and threshold detection, which enables it to determine the target status. Experimental results with the COSMO-SkyMed X-band spotlight SAR images demonstrate the effectiveness of the proposed method.

The remote sensing Synthetic using Aperture Radar can obtain the information of moving targets on the earth surface using MTI (Moving Target Indication) technique, which detects differentiation of the surface image data obtained in successive sessions taken with two antennas mounted on one satellite or two satellites flying in a close formation. In the latter case, the operation of the differentiation is usually performed at the ground station, then each satellite needs to transmit the large amount of data to the ground station separately, which requires broad downlink bandwidths. However, the earth surface image data at the two satellites clearly have a strong correlation since the static background image data other than the moving targets are essentially the same except for some noise factors. Focusing on this fact, we claim that the transmission data can and should be reduced by taking off the redundancy of the two data and sending only the differential parts. In this paper, we discuss the coding theory matter related to this redundancy suppression first, and then discuss source, channel and network coding, and also the Slepian-Wolf theorem for two correlated sources. We next propose a coding and compressing scheme for a certain simplified model. We further generalize it in order to apply this idea to a more realistic model and present a procedure of the proposed scheme.

This paper presents an experimental result to demonstrate azimuth ambiguity suppression with the displaced phase center antenna technique that is promising technique to realize the high resolution wide swath imagery for spaceborne synthetic aperture radar (SAR). The dual receiver antenna data was acquired by airborne Ku-band SAR. In this paper, we propose a signal compensation procedure to cope with beam squint for the following coherent signal processing such as the dual channel reconstruction algorithm. The azimuth ambiguity suppression capability is evaluated by comparing the image reconstructed from dual-receiver channel to the single receiver channel image. The experimental result shows significant azimuth ambiguity suppression.

Along track interferometric (ATI) synthetic aperture radar (SAR) has the capability of detecting moving targets and measuring the line-of-sight velocity of the target. With the ATI SAR, the current velocity map can be obtained by combining two perpendicular ATI data sets recorded over sea in rapid succession. In this paper, we describe how sea surface current can be mapped using airborne Ku-band SAR and present results of sea surface current at Cape Irago in Japan.

Statistical resolution limit (SRL) is a simple metric that provides the highest resolution achievable by any unbiased parametric spectral estimator. In this paper, the SRL of polarimetric radar is derived and compared with that of single polarization radar. The SRL is defined using the Cramer-Rao bound (CRB) of signal parameters for the cases where two point targets are closely located. It is shown analytically and numerically that the polarization information helps improving the SRL.

Passive bistatic radars (PBRs) exploit existing transmitters such as TV broadcasts as the source of illumination. PBR is consisted of two receivers, with one antenna pointed at the source and the other at the target, and the target range is determined by correlating the signal scattered by the target with the signal directly arrived at the receiver. Since PBR does not transmit any waveform, it consumes lower power, and no frequency allocation is required. We have conducted a field experiment exploiting terrestrial digital TV broadcast to assess the feasibility of ISAR (Inverse Synthetic Aperture Radar) observation based on PBR. The wide and flat spectrum of the OFDM signal and the small gap between the channels enables us to obtain high range resolution of up to several meters by exploiting multiple physical channels. Therefore, moderate resolution ISAR imaging is expected to be possible. In this paper, PBR based ISAR algorithm is briefly explained, and the first example of observed PBR based ISAR image of an aircraft target is shown.

Inverse synthetic aperture radar (ISAR) is one of the radar techniques used to observe two-dimensional images of a remotely based target using radio waves. If we keep observing the target and consecutively generate multiple ISAR images, which we call ISAR movie, the target image varies considerably due to the motion of the target. The authors have proposed an algorithm for reconstructing three dimensional target shape from an ISAR movie; however, the algorithm requires a priori knowledge of the relative motion of the target. In this paper, we propose a novel method that estimates the relative motion and the three dimensional shape of the target using multistatic ISAR movie.

Synthetic aperture radar (SAR) images are very useful for many applications. However, SAR images require a great deal of experience for the interpretation, since the appearances are very different from optical images. Additionally, the appearances change significantly by the observational condition, such as frequency and incident angle. We propose a SAR image simulating algorithm developed for the aid of the SAR image analysis and training. The simulator is also intended to support the SAR system configuration design by providing the expected images. The main features of the algorithm are the capability of wide area simulation and the low computational costs. The simulated image of 10 km square area is shown and compared with the real image

Resolution of a radar is limited in range by its bandwidth. One of the various super-resolution techniques for improving its resolution is bandwidth extrapolation (BWE). In the technique, a linear prediction model is fitted to the data, and the model is used to extrapolate the bandwidth. In this paper, the concept of BWE is extended, and a new algorithm called polarimetric bandwidth extrapolation (PBWE) applicable to polarimetric radar data is proposed. It is shown through numerical simulations that utilization of full polarization information allows PBWE to improve the resolution beyond the conventional BWE method. Some results of physical simulation experiment using a W-band polarimetric FMCW radar and corner reflectors are shown to confirm the advantage of PBWE.