基于复杂地面背景序列图像的运动弱小目标检测

分析了复杂地面背景下序列图像检测弱小目标的困难,研究了一种针对场景较复杂的运动弱小目标检测方法。首先将序列图像分割成等间隔的处理单元,并利用最小二乘匹配进行坐标对齐,然后使用Top-Hat算子处理图像并进行CFAR分割,综合考虑处理单元中的每幅图像,对固定背景干扰进行抑制,最后根据运动弱小目标轨迹进行检测。进行了三组序列图像实验并绘制了ROC曲线,证明该方法是一种有效的弱小运动目标检测方法。     

基于星图识别的空间目标检测算法研究

从恒星背景中检测出空间目标是空间目标天基光学测量相机所要解决的关键技术难题之一。提出了一种基于星图识别的空间目标检测算法,利用背景星图的平移、旋转、比例伸缩不变性,应用星图识别原理,根据识别恒星在不同帧图像中的坐标变化估计背景运动参数,进行探测图像序列的配准,然后在配准后的图像序列中检测出空间目标。通过仿真验证,该方法对于运动恒星背景中运动小目标的检测具有比较好的效果。     

基于旋转平台的全空间单目光笔测量方法

针对单目光笔式三维测量系统测量范围小、无法实现全空间测量的问题,提出了一种基于精密旋转台的全空间光笔式单目视觉测量方法。将相机固定在精密旋转台上,在不同的角度对标定板进行拍摄,获取不同角度下相机光心在标定板坐标系下的位置。通过主成分分析(PCA)平面拟合获得相机光心所在平面和相机旋转运动的转轴方向矢量,使用空间最小二乘圆拟合获得相机旋转轴的位置。借助精密转台读数和罗德里格斯公式将转台的旋转角度变换为相机的旋转矩阵和平移向量。利用计算出来的变换矩阵将旋转一定角度后不同位置相机的测量数据转换到同一转台坐标系中,实现全空间测量。实验结果表明,在使用相同光笔的测量系统中,本文方法在基本达到传统单目光笔测量系统测量精度的前提下,可实现360°全空间测量,极大地拓展了单目光笔的应用范围。     

基于多滤波算法融合的红外小目标检测

针对海天背景下红外视频图像序列中小目标检测的精度问题,提出一种基于多滤波融合的算法。通过分析红外小目标及背景噪声的成像特征与时空特性,将Tophat算法与改进的Robinson guard滤波器相结合,有效抑制背景、突显目标;采用自适应阈值分割提取候选目标,并使用Unger平滑滤波以及多目标关联滤波剔除噪声和伪目标。多组红外小目标图像序列的检测实验结果表明,提出的算法平均检测率高达99%,在不同场景下均有较高的检测精度。     

畸变校正与帧差法相结合的运动目标检测

提出了一种图像畸变校正与帧差法相结合的运动目标检测方法。首先利用实验测量和曲线拟合的方法对光学成像系统的畸变进行测量校正,以满足运动目标检测的精度要求;然后将畸变校正方法应用在实时获取的视频图像帧中,利用帧差法实现光学成像系统旋转运动时的运动目标检测。实验表明,该算法可行、抗噪性好,能够满足实时性需求。     

基于聚类的相位相关运动估计方法

针对相机运动引起的图像序列运动的问题,提出了一种基于聚类的相位相关块匹配运动估计算法。利用Harris算子分别在相邻帧图像上检测角点,以参考图像角点为中心选取一个矩形块,将块匹配法与相位相关相结合来计算图像间的运动矢量。最后,对获得的多个块的平移量,进行空间聚类从而选取运动估计比较准确的点。实验结果表明：该算法配准精度能达到亚像素,稳定性较好。     

基于全景环带立体成像系统的深度信息估计

提出了一种基于全景环形透镜(PAL)的全景环带立体成像系统的立体信息提取方法。该成像系统由两组共轴PAL单元构成,其成像圆为两内外相接的圆环,能够实时提取360°水平视场角物体的立体信息。基于该系统的成像原理,针对非单视点成像系统特点建立相机模型对系统进行标定,并验证了标定结果。将尺度不变特征变换(SIFT)算法应用到对应点匹配上,充分利用本系统的共线约束来提高匹配的速度和准确度。应用三角测量原理从捕获的图像中提取有效深度信息。进行立体信息提取实验,给出了结果和误差分析。该系统能有效获取场景深度信息,3m距离范围误差率在±5.2%内,证明了该系统的可行性。     

光速处理的高精度光电混合相关探测图像运动位移

为测量航天遥感相机因姿态不稳定以及各种扰动引起的图像运动,提出了基于光速处理的高精度光电混合相关探测测量方法.利用高速CCD和主CCD对同一目标进行成像,在曝光时间内高速CCD获取序列图像,利用联合变换相关器对所采集的图像序列进行光学运算,测量出相邻序列图像的运动位移.阐述了光学相关方法测量图像运动的原理,并模拟分析了噪声和不同运动条件下的测量精度.建立了相应的测量像移的实验系统,实验数字模拟及实验结果都证实了该方法的有效性,测量精度优于0.1像元,满足卫星遥感相机的使用要求.     

基于粒子群算法的动态热释电目标跟踪

热释电传感器用于红外目标跟踪,可以通过光学调制探测到运动目标的相对方向,而不能确定其准确距离,对此提出一种群智能搜索的动态热释电跟踪方法。传感器在摆动状态下感知红外目标,发现目标时,记录其探测角,通过分析探测角序列与运动目标轨迹离散点在极坐标系下的关系,将目标轨迹的确定转化为极径序列的确定,然后建立关于极径序列的多维粒子群空间,并构造评价粒子优劣的适应度函数,实现了运动目标轨迹的跟踪。该方法扩大了热释电探测范围,提高了目标跟踪精度,通过人体目标直线、曲线运动跟踪实验得出平均定位误差不超过0.4m,表明方法有效。     

基于FPGA的拼接式CMOS线阵相机系统设计

针对单个线阵相机在大视场下传感器边缘失光的问题,提出了在保证分辨率的前提下使用多个相对较低像素的相机拼接,对视场进行分担的方法,设计了一种基于FPGA的拼接式CMOS线阵相机系统。详细介绍了系统的工作原理、硬件构成和各芯片的程序设计,实现了基于FPGA的多相机图像实时校正算法,并将校正后的图像通过高速USB芯片传输到采集软件。实验结果表明,系统在大视场下相比单相机在视场边缘具有更好的成像质量,可应用于多离散目标的检测。     

基于生物视觉特征和视觉心理学的视频显著性检测算法

提出了一种空域和时域相结合的视频显著性检测算法.对单帧图像,受视觉皮层层次化感知特性和Gestalt视觉心理学的启发,提出了一种层次化的静态显著图检测方法.在底层,通过符合生物视觉特性的特征图像（双对立颜色特征及亮度特征图像）的非线性简化模型来合成特征图像,形成多个候选显著区域;在中层,根据矩阵的最小Frobenius-范数（F-范数）性质选取竞争力最强的候选显著区域作为局部显著区域;在高层,利用Gestalt视觉心理学的核心理论,对在中层得到的局部显著区域进行整合,得到具有整体感知的空域显著图.对序列帧图像,基于运动目标在位置、运动幅度和运动方向一致性的假设,对Lucas-Kanade算法检测出的光流点进行二分类,排除噪声点的干扰,并利用光流点的运动幅度来衡量运动目标运动显著性.最后,基于人类视觉对动态信息与静态信息敏感度的差异提出了一种空域和时域显著图融合的通用模型.实验结果表明,该方法能够抑制视频背景中的噪声并且解决了运动目标稀疏等问题,能够较好地从复杂场景中检测出视频中的显著区域.     

Probability analysis of position errors using uncooled IR stereo camera

This paper analyzes the random phenomenon of 3D positions when tracking moving objects using the infrared (IR) stereo camera, and proposes a probability model of 3D positions. The proposed probability model integrates two random error phenomena. One is the pixel quantization error which is caused by discrete sampling pixels in estimating disparity values of stereo camera. The other is the timing jitter which results from the irregular acquisition-timing in the uncooled IR cameras. This paper derives a probability distribution function by combining jitter model with pixel quantization error. To verify the proposed probability function of 3D positions, the experiments on tracking fast moving objects are performed using IR stereo camera system. The 3D depths of moving object are estimated by stereo matching, and be compared with the ground truth obtained by laser scanner system. According to the experiments, the 3D depths of moving object are estimated within the statistically reliable range which is well derived by the proposed probability distribution. It is expected that the proposed probability model of 3D positions can be applied to various IR stereo camera systems that deal with fast moving objects.     

Stereoscopic video surveillance system for detection of target’s 3D location coordinates and moving trajectories

In this paper, a novel stereoscopic video surveillance system to track a target person and detect its three-dimensional (3D) location coordinates and moving trajectory is implemented by using the pan/tilt-controlled stereo camera system. Stereoscopic video images of a target person are captured with stereo camera system and the target face and its location in the left image plane can be detected by use of YCbCr color model and centroid method. Also the target location in the right image plane can be obtained through correlation between the left face image and the right image by using the binary phase extraction joint transform correlator (BPEJTC) algorithm. With these location data of the target face in the left and right image planes the pan/tilt systems can control the stereo camera to track a target person by positioning the target face at the center of the camera’s FOV and making the focusing points of the right and left camera coincided on the target face. And then, with the resultant pan/tilt angles and geometric parameters of the stereo camera system, the target’s real 3D location coordinates and moving trajectory in the world space can be obtained. From some experiments with 900 frames of stereoscopic video image pairs, it is analyzed that the target’s center location after being tracked is kept to be very low error ratio of 0.62%, 2.04% on the average in the x, y directions of the image plane, respectively. Also, the error ratio between the calculated and measured values of the 3D location coordinates of a target person is found to be very low value of 2.1%, 5.7% on the average in the X and Y directions of the real world space, respectively.     

基于目标运动的复眼式双目视觉测距法

昆虫（螳螂）复眼利用目标的动态差异实现立体视觉,具有视场大、计算简单、实时性高等特点,是立体视觉研究的新方向。为实现复眼立体视觉在机器人视觉导航中的应用,根据复眼结构以及信息处理机制,提出并研究了环形光电传感器的目标快速检测与定位方法。首先,搭建了基于60个光电二极管的等6°夹角分布式环形传感器,形成具有360°视场的环形仿生复眼;其次,建立了基于光流原理的运动目标方位角检测模型,采用傅里叶拟合法实现了方位角检测模型的优化,实现了运动目标方位角与目标距离大视野范围内的简单、快速检测。实验结果表明:1）可实现距离375 mm范围内运动速度为30 mm/s的目标方位角实时检测,测量误差在2°范围内;2）基于目标方位角检测模型,可实现双目阵列传感器在300 mm×375 mm视野重叠区域范围内、平均测量误差在10 mm范围内的立体视觉测距。基于光流的运动目标方位角检测模型可用于实现对运动目标空间位置的动态实时检测,在运动检测、视觉导航等领域具有广泛的应用前景。     

Human detection based on motion object extraction and head–shoulder feature

Aimed at the shortcomings of the traditional video monitoring system, human detection method in intelligent video monitoring system was researched. This paper proposed a human detection method based on motion object extraction and head–shoulder feature to complete human detection and statistics in video image sequences. Firstly, background subtraction based on adaptive threshold was used to extract foreground moving object information, then image erosion and image dilation were used to bypass the object shade and remove false object in order to optimize the results of motion object extraction. And finally, for realizing human moving object detection, we proposed the object discrimination algorithm based on human head–shoulder feature to complete human detection and statistics. Experimental results show that the method can successfully realize human detection and statistics. The method is highly accurate and has good real-time and extensive applications. The identification rate is 86% through human video sequences to test. This method can detect human automatically and provide the theoretical and technological base for object detection in the intelligent surveillance system.     

Measurements-based Moving Target Detection in Quantum Video

A method to detect a moving target in multi-channel quantum video is proposed based on multiple measurements on the video strip. The proposed method is capable of detecting the location of the moving target in each frame of the quantum video thereby ensuring that the motion trail of the object is easily and efficiently retrieved. Three experiments, i.e. moving target detection (MTD) of a pixel, MTD of an object in complex shape, and MTD of a pixel whose color is conterminous with that of its background, are implemented to demonstrate the feasibility of the proposal. This study presents a modest attempt to focus on the moving target detection and its applications in quantum video.     

Implementation of opto-digital stereo object tracking system

In this paper, a new opto-digital stereo object tracking system using the variable window mask and the optical binary phase extraction joint transform correlator (BPEJTC) is proposed. At the first step, with the distance information from the stereo camera to the tracking object easily acquired by the structural elements of a stereo vision system, the area of the tracking object can be digitally extracted by using the variable window mask. And, at the second step, by carrying out the optical BPEJTC between this reference image obtained from the variable window mask and the stereo input image, the coordinates of the tracking object's location can be acquired, and then with these values, the convergence angle and the pan/tilt of the stereo tracking camera can be finally controlled. From some experimental results, the proposed system is found to be able to effectively extract the area of the target object from the input image having the background noises by using the variable window mask. And, with the location values of the tracking object obtained by using the optical BPEJTC, the convergence angle and the pan/tilt of the stereo cameras can be controlled. Finally, a feasibility of real-time implementation of the adaptive stereo object tracking system using the proposed algorithm is also suggested.     

用计算机采集、处理摄象机摄取的运动物体信息——提高和改进课堂演示实验的尝试

介绍用计算机采集、处理摄象机摄取的运动物体信息的ＳＨＪＸ系统及经此系统采集和处理的几个物理演示实验中的运动物体的信息     

Vehicle detection and tracking based on phase-correlation

This paper presents vehicle detection and tracking algorithms based on real-time background (RTB) and phase-correlation (PC) in the video sequence of urban highway with fixed camera. Firstly moving objects are obtained by subtracting RTB from serial images. After classification, PC is used to determine corresponding regions of moving objects between consecutive frames. The problems of vehicles' merging and splitting, sudden stop, and restart are also considered. Experiments show that the method is practical and can realize real-time detection and tracking of vehicles on highway.