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。     

基于几何矩阵的增强虚拟现实

增强虚拟现实是组合真实场景和虚拟场景,以达到丰富真实场景的目的.针对摄像机相对景深做较小运动时所获得的视频,提出了一种基于几何矩阵的增强虚拟现实方法.摄像机在这种运动情况下,可以用一个几何矩阵来描述两帧之间的关系,这样就大大简化真实场景和虚拟场景的组合.同时,由于虚拟场景来自二维图像,因此,非常容易对准虚拟物体在视频中的位置.理论和真实视频实验表明,本文提出的增强虚拟现实方法具有计算量小和真实性强等优点.     

单目主动视觉无人机导引中摄像机内参数标定的线性方法

提出一种单目主动视觉无人机导引中摄像机内参数标定的线性方法,在摄像机主点位置进行预标定的条件下,摄像机只需拍摄一幅着舰平面靶标上的一组正方形图像,无须知道该组正方形的任何几何信息,通过计算圆环点,建立绝对二次曲线对摄像机内参数的约束方程即可线性求解摄像机内参数,仿真实验和真实图象实验表明:该方法在摄像机内参数(如焦距等)需要经常改变的视觉任务,如基于视觉的无人机着舰导引、运动参数估计、三维视觉重建及视觉监控中,使得摄像机内参数的标定具有简便、实时和鲁棒性强的特点。     

Maritime border control multisensor system

This focuses on the classification task performed into a multi-sensor system for the coastal surveillance. The system is composed of two platforms of sensors: a land-based platform equipped with a land based radar, an Automatic Identification System (AIS) and an infrared camera (IR); an airborne platform carrying an airborne radar that can operate in a spotlight Synthetic Aperture Radar (SAR) mode, a video camera, and a second IR camera. The tasks performed by the system are the detection, tracking, identification, and classification of multiple targets, the evaluation of their threat level, and the selection of an intervention on them. The classification algorithm implemented inside the system exploits an analytical approach based on the confusion matrix (CM) of the imaging sensors that belong to the system. Some measures of effectiveness (MoE) of the system are evaluated, considering both cases where an ideal error-free classification process and a non-ideal classification process are performed.     

基于三焦距张量及BMA的运动估计和补偿

为克服块匹配位移估计方法 (BMA)不能准确描述物体的旋转、相机镜头的推移 (pan)及缩放的缺点 ,引入计算机视觉中的三线性关系及基于三焦距张量的像素转移进行运动估计和补偿。将像素转移与块匹配运动估计和补偿方法相结合 ,通过实际的图像序列比较了这种方法与单纯 BMA方法的运动补偿效果 ,结果表明 ,在有相机平移及景物有深度变化时 ,本文方法的效果明显优于单纯 BMA方法。     

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     

Automatic CCTV surveillance-towards the VIRTUAL GUARD

The VIRTUAL GUARD is a general-purpose computer-based CCTV surveillance system for detecting potential criminal activity in public areas. The system monitors all activity in the surveillance area, the vast majority of which is people innocently going about their normal business. It will alarm when the observed activities of particular pedestrians and vehicles match any of the pre-defined “suspicious behaviour criteria” programmed into the system. At the same time as analysing movement behaviour, the system uses computer-controlled pan tilt zoom cameras to obtain close-up video recordings of any pedestrians and vehicles at the scene. The system can provide automatic surveillance in many different situations, from parking areas and commercial districts, to housing, recreational and transport facilities. It is particularly suited to the protection of government or commercial buildings located on city streets or other public areas where it is not possible to install perimeter fences     

Method for Optical-Flow-Based Precision Missile Guidance

A new precision guidance law is presented for three-dimensional intercepts against a moving target. In contrast to previously published guidance laws, it does not require knowledge of the range to the target. This makes it appropriate for use on platforms which have an imaging device, such as a video camera, as a primary sensor. We prove that with idealized dynamic model, the guidance law results in zero miss distance, and a formula is given for impact angle error which tends to zero as does target speed, making this method particularly suitable against slow moving targets. Computer simulations are used to test the law with a more realistic model, with a video camera and optical-flow algorithm providing target information. It is shown to perform well compared with another law from the literature, despite requiring less information.     

单目相机物体位姿估计方法研究

现有的机器视觉通常以边缘轮廓和角点作为特征,因此要求背景单一,对环境结构化依赖程度高。为了拓展机器人的应用范围,使其脱离结构化的环境,提出了一种基于SIFT特征点和PNP技术的单目相机估计目标物体位姿的方法。以BumbleBee双目相机为硬件基础,以C++为开发平台,结合了Eigen计算库、OpenCV图像处理库和Triclops库,开发了单目视觉位姿估计算法,实现在复杂背景下对表面纹理较为丰富的物体的位姿估计。利用试验对所提方法进行了验证,试验结果表明,该算法具有较高的估计精度,可以作为机器抓取的依据。     

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.     

基于概率模型运动检测技术的视频对象分割技术研究

MPEG-4最重要的特点是用音频-视频对象来描述内容和进行编码。这就需要按照视频对象进行视频图像分割,而视频对象具有时域和空域这两大特点,而时域分割又占主要的地位。研究了一种时域分割算法,即研究了一种运动检测技术方案,采用Til Aach,Andre Kaup提出的基于概率模型的运动检测技术,用Gauss模型描述噪声引起的像素值变化,将运动目标和有噪声的背景区别开来。最后,采用二值开闭重建滤波器来去除前景和背景中的“小洞”,从而获取了最终的运动检测掩码。     

Acta Aeronautica et Astronautica Sinica

@@ABSTRACTS Motion Estimation and Compensation Based on Trifocal Tensor and BMA.WANG Yang-li, WU Cheng-ke(School of Information Engineering, Xidian University, Xi′an 710071, China).Acta Aeronautica et Astronautica Sinica(ISSN 1000-6893),Vol.22,no.2,Mar.2001,P117-120. 10 refs.in Chinese. To overcome some drawbacks of the block matching algorithm (BMA), which can not accurately describe object rotation, camera panning and zooming, this paper introduces the trilinear constraint and trifocal-tensor-based pixel transfer used in computer vision for motion estimation and compensation. In this method, the trifocal tensor is first estimated by an algorithm based on projection matrix, corresponding relations between two of three images are then determined by BMA and represented with block-based motion vectors, and finally, motion estimation and compensation are carried out by trifocal transferring. As it is not accurate to approximate the dense correspondence between two of the three views by the above block-based motion vectors, using trifocal transferring only will not yield good compensation results. To solve this problem, the paper further puts forward a motion estimation method combining trifocal transferring with BMA. Experiments have been carried out on video sequences to compare the performance of this method with that of BMA, and the results show that this method is superior when there are scene depth variation and camera translation.     

一种基于背景自适应的运动目标检测与跟踪算法

背景图像差分法是运动目标实时检测中常用的方法,但缺乏背景图像随监视场景光照变化而及时更新的合理方法,限制了该方法的适应性。对此,文章首先提出了一种自适应背景更新方法;然后利用最大类间方差法实现运动目标的自适应阈值分割,并利用基于形态学方法的连通区检测算法检测运动目标;最后以Kalman滤波为运动模型实现对运动目标的连续跟踪。实验结果表明:所提方法可随着光照条件的变化,实时、准确地检测出运动目标并实现稳定跟踪。     

大型上单翼飞机机翼三维全场变形测量方案

针对大型上单翼飞机在飞行过程中机翼大挠度变形检测难题,提出了大倾角相机视场下机翼的非接触三维全场变形测量方案。根据上单翼飞机结构特点,将预先标定内参数和相机外参数的共轭相机组安装于飞机垂直尾翼上,采集飞行中的机翼变形图像。首先,提出了大倾角弱相关散斑匹配方法,解决了相机在大倾斜角度状态下采集到的机翼变形弱相关图像相关性差,难以相关匹配的问题。其次,由于测量相机安装于垂直尾翼,飞行测量过程中相机会受到气流扰动产生振动,本文提出了一种相机动态校正方法,通过在机背布置预拉伸刚性不动编码标志点,实时解算基准相机的绝对外参数,进而确定共轭相机的绝对外参数,实现所有测量相机外参数的动态校正。最后,开发了机翼变形全场测量软硬件系统,搭建了缩小比例机翼模型试验台并进行了仿真测量,对系统测量精度进行了比对分析。测量结果验证了本方案的有效性、可行性,对实机测量有一定的指导意义。     

Bayesian Approach to Extended Object and Cluster Tracking using Random Matrices

In algorithms for tracking and sensor data fusion the targets to be observed are usually considered as point source objects; i.e., compared with the sensor resolution their extension is neglected. Due to the increasing resolution capabilities of modern sensors, however, this assumption is often no longer valid as different scattering centers of an object can cause distinct detections when passing the signal processing chain. Examples of extended targets are found in short-range applications (littoral surveillance, autonomous weapons, or robotics). A collectively moving target group can also be considered as an extended target. This point of view is the more appropriate, the smaller the mutual distances between the individual targets are. Due to the resulting data association and resolution conflicts any attempt of tracking the individual objects within the group seems to be no longer reasonable. With simulated sensor data produced by a partly unresolvable aircraft formation the addressed phenomena are illustrated and an approximate Bayesian solution to the resulting tracking problem is proposed. Ellipsoidal object extensions are modeled by random matrices, which are treated as additional state variables to be estimated or tracked. We expect that the resulting tracking algorithms are also relevant for tracking large, collectively moving target swarms.     

结合核相关滤波器和深度学习的运动相机中无人机目标检测

针对无人机与相机快速相对运动造成的运动模糊问题，以及小型无人机外观信息缺失和背景复杂造成漏警和虚警问题，提出了一种新的无人机检测-跟踪方法。针对成像尺寸小于32像素×32像素的无人机目标，提出改进的多层特征金字塔的分类和目标框回归器作为目标检测器，克服漏警。利用检测结果初始化基于核相关滤波的目标跟踪器，并持续修正跟踪结果，跟踪结果为剔除检测器虚警提供依据。在跟踪过程中，引入对观测场景纹理自适应的相机运动补偿策略实现目标重定位。多场景下的实验结果表明：提出的方法在对高速运动小目标的检测和跟踪指标上显著优于传统方法，且运动补偿机制的引入进一步增强了方法在极端复杂场景下的鲁棒性。     

MTI data clustering and formation recognition

Image exploitation technology approaches have generally focused on the detection and spatial analysis of stationary groups of objects on the ground using various sensors. While spatial arrangement is clearly necessary in analyzing military formations, it is usually not sufficient. Typically the arrangement must be examined within some context in order to interpret a pattern of deployment. For moving objects the spatial arrangement of the group relative to the direction of motion is key to recognizing the formation. By examining ground moving target indicator (MTI) radar data over time, motion can be inferred and used to establish a context for interpreting the spatial arrangement of the data. New techniques that exploit the multitemporal nature of MTI data are described. The first is a space-time clustering technique that locates compact groups of objects that persist in time. The technique Is an application of Marr and Hildreth's edge detection methodology to the dual problem of region segmentation, or more accurately, volumetric segmentation of space-time. The second technique is based on the use of the Hough transform for recognizing moving formations such as columns, wedges, and lines abreast by analyzing the shape of clustered MTI detections (specifically the orientation of linear arrangements within the group) with respect to their direction of motion. Preliminary results from simulated MTI data sets are presented     

Real-time drogue recognition and 3D locating for UAV autonomous aerial refueling based on monocular machine vision

Drogue recognition and 3D locating is a key problem during the docking phase of the autonomous aerial refueling(AAR). To solve this problem, a novel and effective method based on monocular vision is presented in this paper. Firstly, by employing computer vision with red-ring-shape feature, a drogue detection and recognition algorithm is proposed to guarantee safety and ensure the robustness to the drogue diversity and the changes in environmental conditions, without using a set of infrared light emitting diodes(LEDs) on the parachute part of the drogue. Secondly, considering camera lens distortion, a monocular vision measurement algorithm for drogue 3D locating is designed to ensure the accuracy and real-time performance of the system,with the drogue attitude provided. Finally, experiments are conducted to demonstrate the effectiveness of the proposed method. Experimental results show the performances of the entire system in contrast with other methods, which validates that the proposed method can recognize and locate the drogue three dimensionally, rapidly and precisely.     

Hypersonic Re-Entry Optical Transfer Equations And Photographic Techniques

Radiation transfer equations applicable to various types of imaging instruments used against distant sources are presented. Emphasis is placed on measurements against point and line radiators made with instruments yielding image spectra. Framing and streaking cameras are discussed in terms of the overall transfer functions of the instrument and sensor. Calibration techniques used for absolute intensity measurements are shown to yield data within a factor of two of the actual value in most cases. The instruments described are of specific use in optical radiometry against bodies penetrating the earth's atmosphere in the hypervelocity regime, and are equally applicable to measurements in a ballistic range. For the problem of tracking these fast-moving objects, imaging instruments are preferred to point detector devices.