Modeling and global trajectory tracking control for an over-actuated MAV

This paper deals with an original micro aerial vehicle (MAV) design, the Omnicopter MAV. It has two central coaxial rotors with fixed-pitch propellers and three perimeter mounted ducted fans with servo motors performing thrust vectoring. Compared with traditional rotary wing MAVs that have inherent underactuation, the Omnicopter possesses some advantages in mobility, for example, lateral translation with zero attitude and hover with nonzero attitude. The trajectory tracking control design, global stability analysis, and control allocation are demonstrated through numerical simulation. The advantage of zero attitude translation is illustrated through experimental results.     

Human tracking system based on adaptive multi-feature mean-shift for robot under the double-layer locating mechanism

Human tracking has been a challenging task for robot in the past decades. In this paper, to realize the human following in a cluttered environment, a human tracking system based on adaptive multi-feature mean-shift (AMF-MS) under the double-layer locating mechanism (DLLM) is proposed to solve the problem of distinguishing target, occlusion, and quick turning. The DLLM, considering the course location processing and fine location processing, is designed to estimate the person’s position using the fusion of heterogeneous data. As an ID tag attached on target can be detected by RF antennas, the course locating method can track the target easily and quickly. The Bayes rule is introduced to calculate the probability where the tag exists due to the instability of RF signals. In the fine locating step, the AMF-MS is proposed because it can reduce computational load and represent target by multi-feature histogram function. Meanwhile, we combine extended Kalman filter and AMF-MS to overcome MS’s inability of occlusion. To control the robot following the target person precisely, an intelligent gear shift strategy based on fuzzy control is implemented by analyzing the robot structure. Experiments demonstrate that the proposed approach is robust to handle complex tracking conditions, and show the system has an optimum performance.     

Online learning and fusion of orientation appearance models for robust rigid object tracking

We introduce a robust framework for learning and fusing of orientation appearance models based on both texture and depth information for rigid object tracking. Our framework fuses data obtained from a standard visual camera and dense depth maps obtained by low-cost consumer depth cameras such as the Kinect. To combine these two completely different modalities, we propose to use features that do not depend on the data representation: angles. More specifically, our framework combines image gradient orientations as extracted from intensity images with the directions of surface normals computed from dense depth fields. We propose to capture the correlations between the obtained orientation appearance models using a fusion approach motivated by the original Active Appearance Models (AAMs). To incorporate these features in a learning framework, we use a robust kernel based on the Euler representation of angles which does not require off-line training, and can be efficiently implemented online. The robustness of learning from orientation appearance models is presented both theoretically and experimentally in this work. This kernel enables us to cope with gross measurement errors, missing data as well as other typical problems such as illumination changes and occlusions. By combining the proposed models with a particle filter, the proposed framework was used for performing 2D plus 3D rigid object tracking, achieving robust performance in very difficult tracking scenarios including extreme pose variations.     

Automatic annotation of tennis games: An integration of audio,vision, and learning

Fully automatic annotation of tennis game using broadcast video is a task with a great potential but with enormous challenges. In this paper we describe our approach to this task, which integrates computer vision, machine listening, and machine learning. At the low level processing, we improve upon our previously proposed state-of-the-art tennis ball tracking algorithm and employ audio signal processing techniques to detect key events and construct features for classifying the events. At high level analysis, we model event classification as a sequence labelling problem, and investigate four machine learning techniques using simulated event sequences. Finally, we evaluate our proposed approach on three real world tennis games, and discuss the interplay between audio, vision and learning. To the best of our knowledge, our system is the only one that can annotate tennis game at such a detailed level.     

结合方向纹理熵的Haar-like特征在线boosting跟踪算法

针对传统基于Haar-like特征的on-line boosting跟踪算法（HBT）需要产生大规模随机特征、占用大量计算资源和存储空间的缺点,提出结合方向纹理熵的Haar-like特征在线boosting跟踪算法（HBTT）。HBTT算法利用灰度共生矩阵的熵获得目标纹理的方向信息,在此基础上有针对性地产生具有方向纹理信息的Haar-like特征,从而可有效避免无效随机特征的产生,减小特征池容量;更进一步,可根据目标纹理的复杂程度自动调整特征数量,使得算法更灵活。在跟踪过程中,在线学习模块可以使错误率较高的特征被结合了目标纹理方向信息的Haar-like特征所替换。与HBT算法比较,HBTT算法的跟踪误差降低了10%以上;在相同特征池容量下,置信度提高了2%以上。实验结果表明,该算法不仅具有较高的鲁棒性,而且在跟踪效率和性能上都有所提高。     

不同重力环境的空间机械臂自抗扰轨迹跟踪控制

针对空间机械臂在地面装调与空间应用时,由于重力环境变化导致机械臂模型发生变化的问题,提出了一种自抗扰控制算法,用于完成空间机械臂轨迹跟踪控制的任务.该算法通过将系统模型及未知外扰作为系统的总和扰动,并利用扩张状态观测器对该扰动进行观测且给予补偿,从而提高了系统抗扰的性能.当机械臂模型随重力环境变化而发生变化时,使用同一个自抗扰控制器对其末端轨迹进行控制,均能取得较好的控制效果.通过对系统的稳定性进行分析,证明了所设计控制器的有效性.将仿真结果与PD控制及自适应鲁棒控制做比较,结果表明该控制算法不仅能适应机械臂模型的变化而且还能有效抵抗系统的扰动,从而使系统具有较强的鲁棒性.     

基于数字式红外传感器的姿态测量盲区补偿法

单轴红外姿态测量系统在测量小型无人机姿态角时存在盲区,通过增加一组与原轴线相垂直的红外传感器,形成双轴姿态测量系统,对盲区进行补偿。两组红外传感器一定有一组位于倾角可测区域,通过测试目标轴的输出温差的大小,判断位于可测区域的轴线,可实现对目标轴姿态角的解算。首次采用数字式输出红外热电堆传感器设计实现了180°无盲区的姿态角测量模块,经过测试,其静态误差小于2°。     

一种基于FPGA的实时视频跟踪系统硬件平台设计

提出一种基于FPGA的视频跟踪系统硬件平台的设计方法,包括主控芯片、CMOS图像采集、多端口SDRAM控制器、图像VGA显示、SOPC系统和云台控制驱动等电路的设计。通过该平台,可以实现实时图像采集、缓存、显示和云台驱动;在SOPC系统中加入代码,可以实现运动目标的检测、跟踪、定位,改变云台转动方向,实现自主跟踪。     

基于Hamiltonian马氏链蒙特卡罗方法的突变运动跟踪

在计算机视觉领域,由镜头切换、目标动力学突变、低帧率视频等引起的突变运动存在极大的不确定性,使得突变运动跟踪成为该领域的挑战性课题.以贝叶斯滤波框架为基础,提出一种基于有序超松弛Hamiltonian马氏链蒙特卡罗方法的突变运动跟踪算法.该算法将Hamiltonian动力学融入MCMC（Markov chain Monte Carlo）算法,目标状态被扩张为原始目标状态变量与一个动量项的组合.在提议阶段,为抑制由Gibbs采样带来的随机游动行为,提出采用有序超松弛迭代方法来抽取目标动量项.同时,提出自适应步长的Hamiltonian动力学实现方法,在跟踪过程中自适应地调整步长,以减少模拟误差.提出的跟踪算法可以避免传统的基于随机游动的MCMC跟踪算法所存在的局部最优问题,提高了跟踪的准确性而不需要额外的计算时间.实验结果表明,该算法在处理多种类型的突变运动时表现出出色的处理能力.     

基于红外传感原理的无人机姿态测量系统设计

由于红外温度传感器能有效感知天空间的热辐射,并且相比传统姿态测量的传感器具有体积小、重量轻、成本低等特点,设计一种基于红外传感原理的无人机姿态测量系统;该系统以新型的ARM Cortex—M3内核微处理器STM32F103ZET6作为处理单元,通过采集红外传感器的姿态信息,进行姿态解算,进而获得无人机的姿态角;同时分析了太阳干扰对实验的影响。实验测试表明:该姿态测量系统简单实用,性能稳定,精度高,能有效满足一般无人机姿态测量的要求。