基于特征块统计的摄像机跟踪算法

提出一种基于特征块统计的摄像机跟踪算法,可用于视频中摄像机运动的快速跟踪和定位。首先在视频范围内抛洒N个随机点,在随机点周围一定区域内寻找出最具颜色差异的像素块作为特征块,然后分析相邻帧中与此特征块最佳匹配的新位置,根据各块的移动情况求出均值,再去掉移动方差过大的特征块,保留余下的特征块进行统计。用最小二乘法统计方法求出连续视频帧之间运动参数的线性变换方程。实验表明,该算法具有较好的跟踪检测效果,对各类视频文件具有较好的鲁棒性,在普通PC机中实现了较为精确的摄像机运动跟踪检测。     

基于块特征分类的运动估计算法

运动估计作为H.264中最重要最耗时的部分,大量的算法研究都是通过减少搜索点数来降低计算量。视频序列中,人们感兴趣的往往是运动的物体,背景部分一般拥有大量的静止块,帧间背景间的运动可以认为完全是由摄像机的运动引起的。利用背景块与前景块的图像特征结合MV预测提出一种新算法。该算法包括以下策略：静止块判断、背景块、前景块和边缘块的检测、起点预测和搜索窗口的改进,同时引进了边缘检测技术。实验结果表明,该算法与全搜索算法(FS)、UMHex、SUMHex相比,搜索时间减少了,而搜索精度接近FS, 码率只有很小的增加。     

Camera movement for chasing a subject with unknown behavior based on real-time viewpoint goodness evaluation

We propose a method that automatically generates a smooth chase camera movement to follow a subject, a user-controlled character or a character with unknown behaviors, in a 3D environment freely in real time. We consider three objectives in generating the smooth-camera movement: to avoid collisions with obstacles, to avoid subject occlusions, and to choose a good viewpoint for looking at the subject. We evaluate the goodness of viewpoints by using a viewpoint entropy map and choose the best viewpoint as the goal position of the camera in real time. Afterwards, we move the camera toward the goal position by following the shortest path, found by the A* algorithm, on a roadmap graph. The resulting camera movement has a high degree of freedom and fulfills the three objectives above. Our method is effective for third-person-view 3D applications in tracking the real-time movement of user-controlled characters in exploring a 3D environment.     

基于相机的转台稳定性检测方法研究

高精密转台常用于高精度测量仪器的性能测试,或作为激光扫描仪的稳定平台使用,而转台的稳定性直接影响仪器的测量精度.提出一种利用数码相机对转台稳定性进行标定的方法,在转台运动时,安装在转台上的相机连续获取相片,基于摄影测量原理,计算相机的空间位置和姿态,从而确定转台的运动轨迹.实验结果表明,该方法可以有效地检测转台运动的稳定性.     

基于运动的摄像机定标方法的综述

该文从运动方式角度将摄像机定标方式分为非限定运动方式和限定运动方式两大类,分别对摄像机各种运动方式对定标技术的影响进行了系统的研究,在应用中应该根据具体情况选择合适的摄像机的运动方式,以满足实际问题的需要。     

Fast Electronic Digital Image Stabilization for Off-Road Navigation

Image stabilization can be used as front-end system for many tasks that require dynamic image analysis, such as navigation and tracking of independently moving objects from a moving platform. We present a fast and robust electronic digital image stabilization system that can handle large image displacements based on a two-dimensional feature-based multi-resolution motion estimation technique. The method tracks a small set of features and estimates the movement of the camera between consecutive frames. Stabilization is achieved by combining all motion from a reference frame and warping the current frame back to the reference. The system has been implemented on parallel pipeline image processing hardware (a Datacube MaxVideo 200) connected to a SUN SPARCstation 20/612 via a VME bus adaptor. Experimental results using video sequences taken from a camera mounted on a vehicle moving on rough terrain show the robustness of the system while running at approximately 20 frames/s.     

基于序的空间金字塔池化网络的人群计数方法

视频中的人群计数在智能监控领域具有重要价值. 由于摄像机透视效果、图像背景、人群密度分布不均匀和行人遮挡等干扰因素的制约, 基于底层特征的传统计数方法准确率较低. 本文提出一种基于序的空间金字塔池化(Rank-based spatial pyramid pooling, RSPP)网络的人群计数方法. 该方法将原图像分成多个具有相同透视范围的子区域并在各个子区域分别取不同尺度的子图像块, 采用基于序的空间金字塔池化网络估计子图像块人数, 然后相加所有子图像块人数得出原图像人数. 提出的图像分块方法有效地消除了摄像机透视效果和人群密度分布不均匀对计数的影响. 提出的基于序的空间金字塔池化不仅能够处理多种尺度的子图像块, 而且解决了传统池化方法易损失大量重要信息和易过拟合的问题. 实验结果表明, 本文方法相比于传统方法具有准确率高和鲁棒性好的优点.     

直线运动摄像机在线动态标定

为实现AS-R智能机器人在运动情况下摄像机在线动态标定,提出一种新的基于粒子滤波的直线运动摄像机标定方法。用状态空间方法描述直线运动摄像机模型,把摄像机内参数和位置运动参数作为状态量,特征点图像坐标作为观测量,根据粒子滤波算法求得摄像机内参数和位置运动参数的最优估计,并用双线程实现整个标定过程。AS-R机器人在直线运动情况下的摄像机在线动态标定实验结果表明：该算法是合理可行的,并且具有很高的标定精度和良好的鲁棒性。该方法适用于各种类型的系统噪声。     

增强室内视觉里程计实用性的方法*

针对现有视觉里程计在实时性、鲁棒性和准确性之间难以协调统一的问题,提出增强视觉里程计实用性的方法.分别运用基于图形处理器的定向加速分割测试特征和旋转感知的二进制鲁棒基元独立特征以及K最邻近加速提取、匹配图像的特征点.根据Kinect有效的深度量程剔除深度误差较大的特征点.求解相机帧间运动时,首先采用高效n点透视快速求解相机帧间运动参数的估计,然后将其作为Levenberg-Marquedt迭代法的初值,优化相机帧间运动参数.在运动参数解计算过程中,使用随机采样一致排除特征外点的干扰.实验表明,文中措施可以提高相机运动轨迹的解算速度,在室内环境下获得的相机运动轨迹更准确,鲁棒性更强,因此适用于室内机器人导航及定位.     

Predictive virtual lane method using relative motions between a vehicle and lanes

We propose a new approach for virtual lane prediction. The main contribution of the proposed method is that the predicted virtual lane can be substituted for lane detection using a camera sensor when the camera image processing fails to detect the lane. The proposed method generates the predicted virtual lane using the relative movement between a vehicle and a lane. To predict the lane, a third-order polynomial function of the longitudinal distance is used as a lane model. Each coefficient of the lane polynomial function at the next sampling time is geometrically calculated using the relative movement of a vehicle, the lanes, the longitudinal velocity and the yaw of the vehicle at the present time. Then, the predictive virtual lane at the next sampling time is obtained without the lane information from the camera sensor at the next sampling time. The proposed method is simple enough that it is suitable for real implementation. The performance of the proposed method was evaluated via experiments with a test vehicle.     

Optimal placement of multiple visual sensors considering space coverage and cost constraints

This paper proposes an optimal camera placement method that analyzes static spatial information in various aspects and calculates priorities of spaces using modeling the moving people pattern and simulation of pedestrian movement. To derive characteristics of space and to cover the space efficiently, an agent-based camera placement method has been developed considering the camera performance as well as the space utility extracted from a path finding algorithm. The simulation shows that the method not only determines the optimal number of cameras, but also coordinates the position and orientation of a camera efficiently considering the installation costs. Experimental results show that our approach achieves a great performance enhancement compared to other existing methods.     

Stable camera position control of unmanned aerial vehicle with three‐degree‐of‐freedom manipulator for visual test of bridge inspection

This paper describes a camera position control with aerial manipulator for visual test of bridge inspection. Our developed unmanned aerial vehicle (UAV) has three‐degree‐of‐freedom (3‐DoF) manipulator on its top to execute visual or hammering test of the inspection. This paper focuses on the visual test. A camera was implemented at the end of the manipulator to acquire images of the narrow space of the bridge such as bearings, which the conventional UAV without the camera‐attached manipulators at its top cannot achieve the fine visual test. For the visual test, it is desirable that the camera is above the body with enough distance between the camera and the body. As obvious, the camera position in the inertial coordinate system is effected by the movement of the body. Therefore we implement the camera position control compensating the body movement into the UAV. As a result of an experiment, it is assessed that the proposed control reduces the position error of the camera comparing the one of the body. The mean position error of the camera is 0.039 m that is 51.4% of the one of the body. Our world‐first study enables to acquire the image of the bearing of the bridge by a camera mounted at the end effector of aerial manipulator fixed on UAV.     

Distance estimation using inertial sensor and vision

A simple distance estimation algorithm using inertial sensors and a mono camera is proposed. Two images of a target are obtained by moving a mono camera. The movement of the camera is estimated using inertial sensors and used as the baseline for the distance estimation. Through experiments, the accuracy of the proposed method is evaluated, where the distance error is 3.9% on average in a few meter ranges.     

基于综合匹配的鸟类视频中运动物体的检测

鄱阳湖是中国最大的湿地。每年秋末冬初,会有成千上万只候鸟来此越冬,该地区的自然生态系统和这些鸟类的活动有着密切的联系,因此,研究该地区的鸟类是了解该地生态系统的一个重要方式。本文采用帧间差分法对该地所拍摄鸟类视频数据进行运动物体检测。由于视频在获取过程中存在摄像机的移动,所以在处理过程中首先需要估计背景的运动并加以补偿。背景运动的估计是在常用的块匹配方法的基础上作了一定的改进,通过多个匹配块的匹配结果来计算运动矢量。为提高正确率,块的匹配在差分图像上进行。最后,利用在鄱阳湖录取的视频对本文所述方法进行了实验,实验结果表明本文所述方法对背景的去除和运动物体的检测的有效性。     

Learned Models for Estimation of Rigid and Articulated Human Motion from Stationary or Moving Camera

We propose an approach for modeling, measurement and tracking of rigid and articulated motion as viewed from a stationary or moving camera. We first propose an approach for learning temporal-flow models from exemplar image sequences. The temporal-flow models are represented as a set of orthogonal temporal-flow bases that are learned using principal component analysis of instantaneous flow measurements. Spatial constraints on the temporal-flow are then incorporated to model the movement of regions of rigid or articulated objects. These spatio-temporal flow models are subsequently used as the basis for simultaneous measurement and tracking of brightness motion in image sequences. Then we address the problem of estimating composite independent object and camera image motions. We employ the spatio-temporal flow models learned through observing typical movements of the object from a stationary camera to decompose image motion into independent object and camera motions. The performance of the algorithms is demonstrated on several long image sequences of rigid and articulated bodies in motion.     

Optimal Representative Blocks for the Efficient Tracking of a Moving Object

Optimal representative blocks are proposed for an efficient tracking of a moving object and it is verified experimentally by using a mobile robot with a pan‐tilt camera. The key idea comes from the fact that when the image size of a moving object is shrunk in an image frame according to the distance between the camera of mobile robot and the moving object, the tracking performance of a moving object can be improved by shrinking the size of representative blocks according to the object image size. Motion estimation using edge detection (ED) and block‐matching algorithm (BMA) are often used in the case of moving object tracking by vision sensors. However, these methods often miss the real‐time vision data since these schemes suffer from the heavy computational load. To overcome this problem and to improve the tracking performance, the optimal representative block that can reduce a lot of data to be computed is defined and optimized by changing the size of the representative block according to the size of object in the image frame. The proposed algorithm is verified experimentally by using a mobile robot with a two degree‐of‐freedom active camera. © 2004 Wiley Periodicals, Inc.     

Persistent people tracking and face capture using a PTZ camera

Pan–tilt–zoom (PTZ) camera is a powerful tool in far-field scenarios. However, most of the current PTZ surveillance systems require manual intervention to move the camera to the desired position. In this paper, we address the problem of persistent people tracking and face capture in uncontrolled scenarios using a single PTZ camera, which could prove most helpful in forensic applications. The system first detects and tracks pedestrians in zoomed-out mode. Then, according to a scheduler, the system selects a person to zoom in. In the zoomed-in mode, we detect a set of face images and solve the face–face association and face–person association problems. The system then zooms back out where tracking is continued as people re-appear in the view. The person–person association module associates the people on the schedule list with the people in the current view. The detected faces are associated with the corresponding people and trajectories. Due to the dynamic nature of our problem, e.g. the field of view of the camera changes because of the pan/tilt/zoom movement of the camera, all of the processes including receiving images from the camera and processing must be done in real time. To the best of our knowledge, the proposed method is the first to address the association of face images to people and trajectories using a single PTZ camera. Extensive experiments in challenging indoor and outdoor uncontrolled conditions demonstrate the effectiveness of the proposed system.     

符合人眼视觉特性的前背景匹配快速算法

针对当前虚拟摄影棚系统前背景匹配算法计算量大、延时长、成本高的缺点,设计并实现了一种快速的符合人眼视觉特性的匹配算法。该算法基于虚拟摄影棚前背景匹配原理及摄像机成像原理,依照人眼视觉系统的视觉惰性、眼球平滑跟随运动及视觉美学相关知识,采用开发桌面控制软件同时控制摄像机与背景的方式来实现。     

Coded exposure HDR light‐field video recording

Capturing exposure sequences to compute high dynamic range (HDR) images causes motion blur in cases of camera movement. This also applies to light‐field cameras: frames rendered from multiple blurred HDR light‐field perspectives are also blurred. While the recording times of exposure sequences cannot be reduced for a single‐sensor camera, we demonstrate how this can be achieved for a camera array. Thus, we decrease capturing time and reduce motion blur for HDR light‐field video recording. Applying a spatio‐temporal exposure pattern while capturing frames with a camera array reduces the overall recording time and enables the estimation of camera movement within one light‐field video frame. By estimating depth maps and local point spread functions (PSFs) from multiple perspectives with the same exposure, regional motion deblurring can be supported. Missing exposures at various perspectives are then interpolated.     

利用多曝光对图像进行动态范围增强

当实际场景的明暗动态范围比电子成像器件所能记录的范围更大时，照片就会有曝光过度或者曝光不足的区域，造成高亮或阴暗处细节的损失。本文提出了一种基于多曝光的图像增强方法。该方法通过对同一场景拍摄一组不同曝光度的图像，依据相机响应曲线性质对不同曝光量的图像分块进行信息量的衡量，选取包含信息最丰富的图块进行拼合，再利用一定的融合函数消除图块间的块效应，得到最终的增强图像。实验证明，该方法简单有效，图像增强后高亮与阴暗处的细节都能很好地呈现在一幅图像中，整体效果自然，对噪点控制较好，无明显光晕现象。