Submit Manuscript
Peer Review
Editor Work2015 Vol. 4, No. 6
Previous Issue
2015, 4(6): 609-620.
Abstract
19798KB
This study examines the complexities of using netted radar to recognize and resolve ballistic midcourse targets. The application of micro-motion feature extraction to ballistic mid-course targets is analyzed, and the current status of application and research on micro-motion feature recognition is concluded for singlefunction radar networks such as low- and high-resolution imaging radar networks. Advantages and disadvantages of these networks are discussed with respect to target recognition. Hybrid-mode radar networks combine low- and high-resolution imaging radar and provide a specific reference frequency that is the basis for ballistic target recognition. Main research trends are discussed for hybrid-mode networks that apply micromotion feature extraction to ballistic mid-course targets.
This study examines the complexities of using netted radar to recognize and resolve ballistic midcourse targets. The application of micro-motion feature extraction to ballistic mid-course targets is analyzed, and the current status of application and research on micro-motion feature recognition is concluded for singlefunction radar networks such as low- and high-resolution imaging radar networks. Advantages and disadvantages of these networks are discussed with respect to target recognition. Hybrid-mode radar networks combine low- and high-resolution imaging radar and provide a specific reference frequency that is the basis for ballistic target recognition. Main research trends are discussed for hybrid-mode networks that apply micromotion feature extraction to ballistic mid-course targets.
2015, 4(6): 621-629.
This paper reports the classification of helicopters, propeller-driven aircraft, and turbojet based on differences in their time-domain modulation periods using a conventional radar system. First, we determine the modulation periods of their time-domain echoes. Then, based on the differences in the time-domain modulation periods, we propose a method for the extraction of time-domain correlation features. Finally, based on the simulated and measured data, via a support vector machine classifier, it is proved that the time-domain correlation features can yield the good classification performance, even with the relatively low pulse repetition frequency, which may induce the ambiguity in Doppler-frequency domain.
This paper reports the classification of helicopters, propeller-driven aircraft, and turbojet based on differences in their time-domain modulation periods using a conventional radar system. First, we determine the modulation periods of their time-domain echoes. Then, based on the differences in the time-domain modulation periods, we propose a method for the extraction of time-domain correlation features. Finally, based on the simulated and measured data, via a support vector machine classifier, it is proved that the time-domain correlation features can yield the good classification performance, even with the relatively low pulse repetition frequency, which may induce the ambiguity in Doppler-frequency domain.
2015, 4(6): 630-638.
As one of the most useful phenomena for separating sea clutter and marine targets, micro-Doppler (m-D) describes the refined motion characteristics of a marine target and helps to improve the abilities of radar detection and recognition. In this study, based on maritime radar, the signal model of a target with micromotion in sea clutter is described. Initially, the definitions of micromotion and m-D are briefly reviewed with a description of their details, and a classification of rigid marine targets that exhibit micromotion is introduced. Then, according to the duration of the observation time, we establish two types of signal models, i.e., in one range unit and across range unit. According to the type of motion, we establish separate signal models for non-uniform translational motion and rotational motion. Finally, the properties of micromotion are analyzed using real radar data, and the effectiveness of the established models is verified.
As one of the most useful phenomena for separating sea clutter and marine targets, micro-Doppler (m-D) describes the refined motion characteristics of a marine target and helps to improve the abilities of radar detection and recognition. In this study, based on maritime radar, the signal model of a target with micromotion in sea clutter is described. Initially, the definitions of micromotion and m-D are briefly reviewed with a description of their details, and a classification of rigid marine targets that exhibit micromotion is introduced. Then, according to the duration of the observation time, we establish two types of signal models, i.e., in one range unit and across range unit. According to the type of motion, we establish separate signal models for non-uniform translational motion and rotational motion. Finally, the properties of micromotion are analyzed using real radar data, and the effectiveness of the established models is verified.
2015, 4(6): 639-647.
Radar Automatic Target Recognition (RATR) is the key technique to be breaked through in the fuure development of intelligent weapon system. Compared to the 2-D SAR image target recognition, High Resolution Range Profile (HRRP) target recognition has the advantage of low data dimension, low requirement of radar system's calculation and storage ability, and the imaging algorithm is also not complicated. HRRP imaging is the first and the key process in target recognition, its speed and imaging quality can directly influence the real-time capability and accuracy of target recognition. In this paper a new HRRP imaging algorithm NUFFT algorithm is proposed, the derivation of mathematical expression is given, both for the echo simulation process and the imaging process. In the meantime, by analyzing each step's calculation complexity, we compared the calculation complexity of four different imaging algorithms, we also simulate two target's imaging and target recognition processing. Theoretical analysis and simulation both prove that the proposed algorithm's calculation complexity is improved in various degree compared with the others, thus can be effectively used in target recognition.
Radar Automatic Target Recognition (RATR) is the key technique to be breaked through in the fuure development of intelligent weapon system. Compared to the 2-D SAR image target recognition, High Resolution Range Profile (HRRP) target recognition has the advantage of low data dimension, low requirement of radar system's calculation and storage ability, and the imaging algorithm is also not complicated. HRRP imaging is the first and the key process in target recognition, its speed and imaging quality can directly influence the real-time capability and accuracy of target recognition. In this paper a new HRRP imaging algorithm NUFFT algorithm is proposed, the derivation of mathematical expression is given, both for the echo simulation process and the imaging process. In the meantime, by analyzing each step's calculation complexity, we compared the calculation complexity of four different imaging algorithms, we also simulate two target's imaging and target recognition processing. Theoretical analysis and simulation both prove that the proposed algorithm's calculation complexity is improved in various degree compared with the others, thus can be effectively used in target recognition.
2015, 4(6): 648-657.
One of the unique characteristics of a ground target is its micro-motion, which can be used for target classification and identification. In this study, methods for vibrating ground target detection and feature extraction of the one-stationary bistatic frequency-modulated continuous-wave Synthetic Aperture Radar (SAR) are studied. The Displaced Phase Center Antenna (DPCA) technique is adopted to suppress the ground clutter, allowing the ground-vibrating targets to be detected. Analysis of the received signal indicates that the DPCA processing results in a slow time-varying envelope, known as the Slow Time Envelope (STE). The STE has a direct effect on the micro-Doppler time-frequency curve, which therefore cannot be obtained unbroken. Furthermore, vibrating features are extracted by utilizing their relationship with the STE term. Finally, some simulations are provided to validate the theoretical derivation and effectiveness of the proposed extraction method.
One of the unique characteristics of a ground target is its micro-motion, which can be used for target classification and identification. In this study, methods for vibrating ground target detection and feature extraction of the one-stationary bistatic frequency-modulated continuous-wave Synthetic Aperture Radar (SAR) are studied. The Displaced Phase Center Antenna (DPCA) technique is adopted to suppress the ground clutter, allowing the ground-vibrating targets to be detected. Analysis of the received signal indicates that the DPCA processing results in a slow time-varying envelope, known as the Slow Time Envelope (STE). The STE has a direct effect on the micro-Doppler time-frequency curve, which therefore cannot be obtained unbroken. Furthermore, vibrating features are extracted by utilizing their relationship with the STE term. Finally, some simulations are provided to validate the theoretical derivation and effectiveness of the proposed extraction method.
2015, 4(6): 658-665.
This paper presents a novel texture feature extraction method based on a Gabor filter and Three-Patch Local Binary Patterns (TPLBP) for Synthetic Aperture Rader (SAR) target recognition. First, SAR images are processed by a Gabor filter in different directions to enhance the significant features of the targets and their shadows. Then, the effective local texture features based on the Gabor filtered images are extracted by TPLBP. This not only overcomes the shortcoming of Local Binary Patterns (LBP), which cannot describe texture features for large scale neighborhoods, but also maintains the rotation invariant characteristic which alleviates the impact of the direction variations of SAR targets on recognition performance. Finally, we use an Extreme Learning Machine (ELM) classifier and extract the texture features. The experimental results of MSTAR database demonstrate the effectiveness of the proposed method.
This paper presents a novel texture feature extraction method based on a Gabor filter and Three-Patch Local Binary Patterns (TPLBP) for Synthetic Aperture Rader (SAR) target recognition. First, SAR images are processed by a Gabor filter in different directions to enhance the significant features of the targets and their shadows. Then, the effective local texture features based on the Gabor filtered images are extracted by TPLBP. This not only overcomes the shortcoming of Local Binary Patterns (LBP), which cannot describe texture features for large scale neighborhoods, but also maintains the rotation invariant characteristic which alleviates the impact of the direction variations of SAR targets on recognition performance. Finally, we use an Extreme Learning Machine (ELM) classifier and extract the texture features. The experimental results of MSTAR database demonstrate the effectiveness of the proposed method.
2015, 4(6): 666-673.
This study proposes a model-based Synthetic Aperture Radar (SAR) automatic target recognition algorithm. Scattering is computed offline using the laboratory-developed Bidirectional Analytic Ray Tracing software and the same system parameter settings as the Moving and Stationary Target Acquisition and Recognition (MSTAR) datasets. SAR images are then created by simulated electromagnetic scattering data. Shape features are extracted from the measured and simulated images, and then, matches are searched. The algorithm is verified using three types of targets from MSTAR data and simulated SAR images, and it is shown that the proposed approach is fast and easy to implement with high accuracy.
This study proposes a model-based Synthetic Aperture Radar (SAR) automatic target recognition algorithm. Scattering is computed offline using the laboratory-developed Bidirectional Analytic Ray Tracing software and the same system parameter settings as the Moving and Stationary Target Acquisition and Recognition (MSTAR) datasets. SAR images are then created by simulated electromagnetic scattering data. Shape features are extracted from the measured and simulated images, and then, matches are searched. The algorithm is verified using three types of targets from MSTAR data and simulated SAR images, and it is shown that the proposed approach is fast and easy to implement with high accuracy.
2015, 4(6): 674-680.
We have proposed an improved Sparsity Preserving Projection (SPP) method to implement target feature extraction. It combines the SPP feature extraction using the idea of the Locality Preserving Projection (LPP) scheme to build a new objective function, which can not only maintain the relationship of sparse reconstruction between the samples but also minimize the distance between similar sample types in the projection space. Experimental results with Moving and Stationary Target Acquisition and Recognition (MSTAR) Synthetic Aperture Radar (SAR) data sets show that the average recognition rate using the proposed method is up to 97.81% without knowing the target to be azimuth, which can improve the target recognition result even further for obvious reasons. The proposed method is an effective one for SAR target recognition.
We have proposed an improved Sparsity Preserving Projection (SPP) method to implement target feature extraction. It combines the SPP feature extraction using the idea of the Locality Preserving Projection (LPP) scheme to build a new objective function, which can not only maintain the relationship of sparse reconstruction between the samples but also minimize the distance between similar sample types in the projection space. Experimental results with Moving and Stationary Target Acquisition and Recognition (MSTAR) Synthetic Aperture Radar (SAR) data sets show that the average recognition rate using the proposed method is up to 97.81% without knowing the target to be azimuth, which can improve the target recognition result even further for obvious reasons. The proposed method is an effective one for SAR target recognition.
2015, 4(6): 681-688.
Residual motion error is common in high-resolution circular Synthetic Aperture Radar (SAR) image defocusing. The Signal-to-Clutter Ratio (SCR) in echo data domain is relatively low; thus, the phase error spans several range bins. To solve this problem, we propose a focusing algorithm for circular SAR based on phase-error estimation in the image domain. The method estimates the point-target image window interception and then the phase error from echo regeneration in the defocused image. Subsequently, the range migration error is calculated, and finally, the phase error in the echo data is compensated for azimuth focusing and range cell migration correction. Simulation and real-data processing verified the proposed method.
Residual motion error is common in high-resolution circular Synthetic Aperture Radar (SAR) image defocusing. The Signal-to-Clutter Ratio (SCR) in echo data domain is relatively low; thus, the phase error spans several range bins. To solve this problem, we propose a focusing algorithm for circular SAR based on phase-error estimation in the image domain. The method estimates the point-target image window interception and then the phase error from echo regeneration in the defocused image. Subsequently, the range migration error is calculated, and finally, the phase error in the echo data is compensated for azimuth focusing and range cell migration correction. Simulation and real-data processing verified the proposed method.
2015, 4(6): 689-697.
Interferometric Synthetic Aperture Radar (InSAR) coherence is important not only in determining measurement quality but also for extracting thematic information about objects on the ground in combination with backscattering coefficient. The decorrelation of repeat-pass InSAR caused by soil moisture change has received little attention in comparison with other sources of decorrelation. In this paper, we use ENVISAT ASAR data and laboratory experiments to analyze the repeat-pass InSAR coherence loss results due to soil moisture change. C-band ASAR data has high coherence over bare soil and grassland areas, which indicates that these two types of land cover are good choices for the analysis of InSAR coherence loss due to soil moisture change. In addition, spaceborne SAR with short revisit capability, has great potential for this specific application, even for agricultural fields. We conducted further analysis of the soil-sample laboratory data acquired in an anechoic chamber because of its controllable environment and the ability to exclude other sources of decorrelation. We found that the lower frequency range, 2-2.5 GHz, has the highest coherence and is the most insensitive to the initial soil moisture value. This indicates that the S band is more advantageous than the C band when using InSAR coherence to detect soil moisture change. This is true at least with respect to the S band's high coherence level and insensitivity to initial soil moisture values.
Interferometric Synthetic Aperture Radar (InSAR) coherence is important not only in determining measurement quality but also for extracting thematic information about objects on the ground in combination with backscattering coefficient. The decorrelation of repeat-pass InSAR caused by soil moisture change has received little attention in comparison with other sources of decorrelation. In this paper, we use ENVISAT ASAR data and laboratory experiments to analyze the repeat-pass InSAR coherence loss results due to soil moisture change. C-band ASAR data has high coherence over bare soil and grassland areas, which indicates that these two types of land cover are good choices for the analysis of InSAR coherence loss due to soil moisture change. In addition, spaceborne SAR with short revisit capability, has great potential for this specific application, even for agricultural fields. We conducted further analysis of the soil-sample laboratory data acquired in an anechoic chamber because of its controllable environment and the ability to exclude other sources of decorrelation. We found that the lower frequency range, 2-2.5 GHz, has the highest coherence and is the most insensitive to the initial soil moisture value. This indicates that the S band is more advantageous than the C band when using InSAR coherence to detect soil moisture change. This is true at least with respect to the S band's high coherence level and insensitivity to initial soil moisture values.
2015, 4(6): 698-707.
The Circular Synthetic Aperture Radar (CSAR) can obtain the entire scattering properties of targets because of its great ability of 360 observation. In this study, an optimal orientation of the CSAR imaging algorithm of buildings is proposed by applying a combination of coherent and incoherent processing techniques. FEKO software is used to construct the electromagnetic scattering modes and simulate the radar echo. The FEKO imaging results are compared with the isotropic scattering results. On comparison, the optimal azimuth coherent accumulation angle of CSAR imaging of buildings is obtained. Practically, the scattering directions of buildings are unknown; therefore, we divide the 360 echo of CSAR into many overlapped and few angle echoes corresponding to the sub-aperture and then perform an imaging procedure on each sub-aperture. Sub-aperture imaging results are applied to obtain the all-around image using incoherent fusion techniques. The polarimetry decomposition method is used to decompose the all-around image and further retrieve the edge information of buildings successfully. The proposed method is validated with P-band airborne CSAR data from Sichuan, China.
The Circular Synthetic Aperture Radar (CSAR) can obtain the entire scattering properties of targets because of its great ability of 360 observation. In this study, an optimal orientation of the CSAR imaging algorithm of buildings is proposed by applying a combination of coherent and incoherent processing techniques. FEKO software is used to construct the electromagnetic scattering modes and simulate the radar echo. The FEKO imaging results are compared with the isotropic scattering results. On comparison, the optimal azimuth coherent accumulation angle of CSAR imaging of buildings is obtained. Practically, the scattering directions of buildings are unknown; therefore, we divide the 360 echo of CSAR into many overlapped and few angle echoes corresponding to the sub-aperture and then perform an imaging procedure on each sub-aperture. Sub-aperture imaging results are applied to obtain the all-around image using incoherent fusion techniques. The polarimetry decomposition method is used to decompose the all-around image and further retrieve the edge information of buildings successfully. The proposed method is validated with P-band airborne CSAR data from Sichuan, China.
2015, 4(6): 708-714.
Circular Synthetic Aperture Radar (CSAR) is a recently developed all-directional high-resolution imaging mode, which is efficient in dealing with the target recognition, area monitoring, and three-dimensional reconstruction because of the acquisition of 360 data in a single pass. To obtain the entire 360 image using information from all subapertures, the following methods are used: (1) the coherent addition method, (2) the incoherent addition method, and (3) the maximum-intensity methods. In this study, different statistical models of speckle in CSAR images are proposed and the speckle reduction of each model is discussed. Experiments show that the incoherent addition and maximum-intensity methods reduce speckle, whereas the coherent addition method does not.
Circular Synthetic Aperture Radar (CSAR) is a recently developed all-directional high-resolution imaging mode, which is efficient in dealing with the target recognition, area monitoring, and three-dimensional reconstruction because of the acquisition of 360 data in a single pass. To obtain the entire 360 image using information from all subapertures, the following methods are used: (1) the coherent addition method, (2) the incoherent addition method, and (3) the maximum-intensity methods. In this study, different statistical models of speckle in CSAR images are proposed and the speckle reduction of each model is discussed. Experiments show that the incoherent addition and maximum-intensity methods reduce speckle, whereas the coherent addition method does not.
微信 | 公众平台 
