无人机载多传感器图像融合评述

为了能为作战指挥系统提供清晰的局部战场信息,提高对局部战场低可观测目标的检测概率、定位精度及识别概率,迫切需要对无人机载SAR、可见光传感器、红外探测器等图像信息及其他非图像信息融合处理。提出了无人机载多传感器图像融合技术需要研究的内容,如: 图像融合新方法研究;无人机载SAR图像的非平稳性处理;基于图像融合目标检测和处理技术; 无人机载图像和非图像信息的融合问题;无人机载多传感器图像融合的实现及评估;多无人机载合成孔径雷达的协同成像;用图像融合的方法实现对运动目标检测等。分析了所提出研究内容的可行性,剖析了其中的关键技术,拟定了可能的技术路线。     

Application of Three-Dimensional Filtering to Moving Target Detection

The standard approach to the detection of a stationary target immersed within an optically observed scene is to use integration to separate the target energy from the background clutter. When the target is nonstationary and moves with fixed velocity relative to the clutter, the procedure for integrating the target signal is no longer obvious. In this paper it is shown that the problem of tracking a target having a fixed velocity can be cast into a general framework of three-dimensional filter theory. From this point of view, the target detection problem reduces to the problem of finding optimal three-dimensional filters in the three-dimensional transform domain and processing the observed scene via this filtering. The design of these filters is presented, taking into account the target, clutter, and optical detection models. Performance is computed for a basic clutter model, showing the effective increase in detectability as a function of the target velocity. The three-dimensional transform approach is readily compatible with VLSI array processing technology.     

A target cueing and tracking system (TCATS) for smart videoprocessing

A smart system that uses state-of-the-art image processing techniques to detect and track targets moving in video or infrared imagery is presented. It is a general-purpose system that was developed to be independent of the application and the user interface. TCATS has successfully demonstrated a high detection capability and the ability to ignore exterior environmental conditions, such as cloud shadows moving through the scene, snow, blowing dust, rain and moderate camera motion. TCATS can handle many targets and retain the identity of each object from one frame to the next. Special rugged hardware has been developed to reduce the size, cost, and power consumption and to increase the processing rate, reliability, and environmental operating capability of the system. TCATS can be incorporated into an existing video assessment system without a major cost outlay intruder. The TCATS algorithms, custom-developed hardware, and preliminary field test results are described     

基于力场转换理论的图像粗大边缘检测方法

基于粗大边缘的异源图像匹配在导航制导等领域具有广阔的应用前景,但是现有边缘检测方法很难提取出异源图像中的粗大边缘.根据异源图像成像原理和灰度分布特点,提出一种基于力场转换理论的异源图像粗大边缘检测新方法.首先,根据引力概念计算图像中各像素点受到合力的大小和方向;其次,为了去除光照和异源图像灰度不同的影响,对图像中像素点...     

A Detection Algorithm for Optical Targets in Clutter

There is active interest in the development of algorithms for detecting weak stationary optical and IR targets in a heavy opticalclutter background. Often only poor detectability of low signal-to-noise ratio (SNR) targets is achieved when the direct correlation method is used. In many cases, this is partly obviated by using detection with correlated reference scenes [1, 2].This paper uses the experimentally justified assumption that most optical clutter can be modeled as a whitened Gaussian randomprocess with a rapidly space-varying mean and a more slowlyvarying covariance [2]. With this assumption, a new constant falsealarm rate (CFAR) detector is developed as an application of the classical generalized maximum likelihood ratio test of Neyman and Pearson. The final CFAR test is a dimensionless ratio. This test exhibits the desirable property that its probability of a false alarm(PFA) is independent of the covariance matrix of the actual noiseencountered. When the underlying noise processes are complex intime, similar considerations can yield a sidelobe canceler CFARdetection criterion for radar and communications. Performance analyses based on the probability of detection (PD)versus signal-to-noise ratio for several given fixed false alarm probabilities are presented. Finally these performance curves are validated by computer simulations of the detection process which use real image data with artificially implanted signals.     

Efficient target tracking using dynamic programming

A dynamic programming (DP) algorithm has been developed for the detection and tracking of subpixel-sized, low signal-to-noise ratio (SNR) targets observed by side-or forward-looking imaging sensors. A distinguishing feature of this approach is that target detection and tracking are combined into a single optimization procedure that takes into account statistical models of target motion, background noise, and clutter. Current work has led to a number of technical innovations that improve the performance and efficiency of the DP tracking algorithm, including the development of a new track scoring function, and an extension to the basic DP algorithm that reduces computation requirements by over an order of magnitude. A prototype infrared (IR) target tracking system incorporating these enhancements has been implemented for a step-starting IR camera application. Sensitivity improvements of several decibels over conventional sequential detection and tracking algorithms were realized     

Radar Image Processing with Clusters of Computers

Some radar image processing algorithms such as shape-from-shading are particularly compute-intensive and time-consuming. If, in addition, a data set to be processed is large, then it may make sense to perform the processing of images on multiple workstations or parallel processing systems. We have implemented shape-from-shading, stereo matching, resampling, gridding, and visualization of terrain models in such a manner that they execute either on parallel machines or on clusters of workstations. We were motivated by the large image data set from NASA's Magellan mission to planet Venus, but received additional inspiration from the European Union's Center for Earth Observation program (CEO) and Austria's MISSION initiative for distributed processing of remote sensing images on remote workstations, using publicly-accessible algorithms. We developed a multi-processor approach that we denote as CDIP for Concurrent and Distributed Image Processing. The speedup for image processing tasks increases nearly linearly with the number of processors, be they on a parallel machine or arranged in a cluster of distributed workstations. Our approach adds benefits for users of complex image processing algorithms: the efforts for code porting and code maintenance are reduced and the necessity for specialized parallel processing hardware is eliminated.     

Detection of Target Maneuver Onset

A classical maneuvering target tracking (MTT) problem (detection of the onset of a target maneuver) is presented in two parts. The first part reviews most traditional maneuver onset detectors and presents results from a comprehensive simulation study and comparison of their performance. Six algorithms for maneuver onset detection are examined: measurement residual chi-square, input estimate chi-square, input estimate significance test, generalized likelihood ratio (GLR), cumulative sum, and marginalized likelihood ratio (MLR) detectors. The second part proposes two novel maneuver onset detectors based on sequential statistical tests. Cumulative sums (CUSUM) type and Shiryayev sequential probability ratio (SSPRT) maneuver onset detectors are developed by using a likelihood marginalization technique to cope with the difficulty that the target maneuver accelerations are unknown. The proposed technique gives explicit solutions for Gaussian-mixture prior distributions, and can be applied to arbitrary prior distributions through Gaussian-mixture approximations. The approach essentially utilizes a~priori information about the maneuver accelerations in typical tracking engagements and thus allows to improve detection performance as compared with traditional maneuver detectors. Simulation results demonstrating the improved capabilities of the proposed onset maneuver detectors are presented.     

A sequential detection approach to target tracking

Sequential hypothesis testing is investigated for multiframe detection and tracking of low-observable maneuvering point-source targets in a digital image sequence. The proposed multiple multistage hypothesis test tracking (MMHTT) algorithm extends tracks formed from sequentially detected target trajectory segments using a multiple hypothesis tracking strategy. The MMHTT algorithm does not require a probabilistic larger maneuver model. Computational efficiency is achieved by using a truncated sequential probability ratio test (SPRT) to prune a dense tree of candidate target trajectories and score the detected trajectory segments. An analytical performance evaluation is presented and confirmed by experimental results from an optical satellite tracking application     

基于局部单应性矩阵的图像拼接与定位算法研究

针对单个摄像机的视野范围有限导致在大场景下监控效果不够理想的问题,提出了一种改进的图像拼接与目标定位算法。该算法以多个摄像机获取的具有共视区域的监控图像为基础,通过对图像进行网格划分后分别计算多个局部单应性矩阵完成初步对准,然后对网格顶点进行微调优化完成最后配准。最后对图像进行融合形成无缝、自然的大视角图像,并利用场景信息在获取的全景图像上对目标进行快速定位,以满足监控人员对场景中目标的全景捕捉分析功能。实验结果表明,该算法能显著提高大场景下图像拼接结果的质量并实现目标的快速定位。     

Super resolution SAR imaging via parametric spectral estimationmethods

Super resolution synthetic aperture radar (SAR) image formation via sophisticated parametric spectral estimation algorithms is considered. Parametric spectral estimation methods are devised based on parametric data models and are used to estimate the model parameters. Since SAR images rather than model parameters are often used in SAR applications, we use the parameter estimates obtained with the parametric methods to simulate data matrices of large dimensions and then use the fast Fourier transform (FFT) methods on them to generate SAR images with super resolution. Experimental examples using the MSTAR and Environmental Research Institute of Michigan (ERIM) data illustrate that robust spectral estimation algorithms can generate SAR images of higher resolution than the conventional FFT methods and enhance the dominant target features     

Radar Target Detection and Map-Matching Algorithm Studies

Results of a study of adaptive threshold target detection and map-matching algorithms are presented. Log threshold processing is shown to be preferred over linear threshold processing when the clutter data surrounding the target cell is contaminated by other targets, decoy corner reflectors, or bright clutter cells. Whereas previous studies have resorted to extensive Monte-Carlo simulations of log threshold algorithms, the results were obtained using a novel analytical approach based upon Parseval's theorem.     

基于区域的卷积神经网络在空对地车辆检测中的应用

针对传统空对地车辆检测算法在光照变换、场景变化时检测效果不佳的问题,提出了基于区域卷积神经网络Faster RCNN模型的空对地车辆检测方法,介绍了Faster RCNN模型以及模型训练过程.实验结果表明,基于Faster RCNN的空对地车辆检测方法是可行的,对不同光照和场景下的车辆检测可以取得较好的效果.     

基于液晶光阀的动态红外场景仿真研究

介绍了基于液晶光阀的动态红外场景仿真系统构成,分析了液晶光阀的波长变换和空间调制机理。对反映红外场景仿真效果主要因素(光谱输出特性、时间响应特性和空间分辨力)进行了理论分析和研究。利用优化设计液晶光阀参数构建的动态红外场景仿真样机,测量其光谱输出特性、时间响应特性和空间分辨力,结果表明,仿真系统的输出光谱可覆盖中波和长波红外波段,可实现帧频70Hz,系统空间分辨力优于10 lp/mm(受红外光学系统和红外热像仪分辨力限制)。     

低信噪比图像目标检测技术研究——Ⅱ．优化聚类与特征识别

作为自适应图像目标检测技术层次化数据处理流程的中层聚类和末端识别模块,本文对其进行了较为深入的探讨,给出了具体应用的几种实现算法。研究了保持形状特性的多结构元组合形态滤波算法,并探讨了采用外推方式挖掘残缺信号特征的应用算法。     

Probability of Detecting a Swerling I Targetet on Two Correlated Observations

Modern radars characterized by electronically steered beams, frequency agility, and range-ambiguous waveforms can use a processing technique whereby each new detection is followed by a second dwell to verify the initial detection. The second dwell is used to minimize false alarms and to resolve range and/or Doppler ambiguities. Under the assumption of a Swerling I target model, the target cross section remains constant when both dwells occur on the same scan and the same transmission frequency is used. Analytic expressions have been developed for the probability of detecting a Swerling I target on both of the correlated dwells (same target cross section). These expressions are used to calculate the probability of a second dwell detection given a first dwell detection and the probability of at least one detection on two dwells. These probabilities are compared with those of independent dwells (independent target cross sections) which result when two transmission frequencies of sufficient separation are utilized.     

A Model for Generating Synthetic VHF SAR Forest Clutter Images

We propose a model for generating low-frequency synthetic aperture radar (SAR) clutter that relates model parameters to physical characteristics of the scene. The model includes both distributed scattering and large-amplitude discrete clutter responses. The model also incorporates the SAR imaging process, which introduces correlation among image pixels. The model may be used to generate synthetic clutter for a range of environmental operating conditions for use in target detection performance evaluation of the radar and automatic target detection/recognition algorithms. We derive a statistical representation of the proposed clutter model's pixel amplitudes and compare with measured data from the CARABAS-II SAR. Simulated clutter images capture the structure and amplitude responses seen in the measured data. A statistical analysis shows an order of magnitude improvement in model fit error compared with standard maximum-likelihood (ML) density fitting methods.     

Preventing Damage by Hidden Objects in Vegetation

The paper presents the experimental radar-based sensor to operate in environment with vegetation cover for detection and discrimination there the small-sized invisible dielectric and metallic objects. This sensor constitutes a vehicle-housed emergency system for surveillance of area with vegetation ahead on the path of moving vehicle to prevent its contact with hidden objects. The emergency system functioning in basically implemented by real-time electromagnetic imaging of the scene of interest and its following image processing to enhance the target responses. The developed and tested experimental radar techniques are under consideration. The results of experimental examinations in field are presented and discussed.     

Optical moving target detection with 3-D matched filtering

Three-dimensional (3-D) matched filtering has been suggested as a powerful processing technique for detecting weak, moving optical targets immersed in a background noise field. The procedure requires the processing of entire sequences of frames of optical scenes containing the moving targets. The 3-D processor must be properly matched to the target signature and its velocity vector, but will simultaneously detect all targets to which it is matched. The results of a study to evaluate the 3-D processor are presented. Simulation results are reported which show the ability of the processor to detect targets well below the background level. These results demonstrate the capability and robustness of the processor, and show that the algorithms, although somewhat complicated, can be implemented readily. Some effects on the number of frames processed, target flight scenarios, and velocity and signature mismatch are also presented. The ability to detect multiple targets is demonstrated     

基于电阻阵列的红外场景生成技术

针对红外成像制导半实物仿真试验的需求,研究了红外场景生成技术。主要通过对场景的建模、目标及场景的红外特性分析,开展了基于可见光图像调制生成红外图像技术的研究。同时对飞机、地面背景及红外诱饵干扰的建模技术进行了探讨,完善了红外场景的生成方法。利用该红外场景生成技术为基于电阻阵列的红外图像转换器提供输入信息,驱动电阻阵列产生红外辐射,经热像仪观测,生成的图像细腻、真实度高,为制导半实物仿真试验提供了更为逼真的红外仿真环境。