基于视频识别的乒乓球发球裁判系统实验研究——针对抛球高度和抛球角度问题

乒乓球作为竞技运动中一类很快速的运动,其技术合法性需要裁判在几秒钟内根据观察和经验做出判断,这是一个很复杂的任务。由于裁判主观判断的不确定及不精确性,会导致比赛中产生一些误判现象,从而对比赛结果产生很大影响。在目前的比赛当中,争议最大的判罚即为运动员发球的抛球高度和抛球角度的合法性问题。为了解决这个争议问题,设计了一种辅助性的基于视频识别的乒乓球发球裁判系统,可以识别并跟踪视频中乒乓球的位置,并准确地确定乒乓球离开执拍手的瞬间。作为一种新型的裁判系统,第一次将机器学习的方法引入乒乓球识别中,并讨论了一种设置预测感兴趣区域的方法来降低识别错误率,提高识别效率。系统通过视频处理最终将得到运动员发球的抛球高度和抛球角度,力图在训练中为运动员提供一定技术参考,在比赛中为裁判员的判罚提供一定的科学参考,也为运动员质疑裁判的判罚提供一种辅助性的依据。     

视频同步极性识别处理电路的设计与实现

在分量视频信号中,通常出于简化系统等目的需要将水平和垂直同步信号合成单路的复合同步信号,然后再和视频信号进行叠加形成复合视频信号。而后续的对复合视频的相关处理中,又要求前面的复合环节中的两个同步信号的极性必须相同,否则无法识别同步信号和视频信号。由此,提出一种有效的同步信号极性识别和复合处理算法,并进行了电路实现。实验证明,采用所设计的电路完全可以解决不同极性同步信号的识别与复合,电路实现简单、可靠。     

基于Lab特征模型的兵乓球追踪机器人的设计与实现

随着科技的进步,机器人是未来发展的趋势。兵乓球追踪机器人是一种典型的实时机器人,追踪是捡球机器人前期的必要基础。针对单目摄像头的运用,提出了一种基于Lab特征模型的兵乓球追踪机器人的设计方法。乒乓球追踪机器人采用图像处理技术和光学理论对乒乓球进行识别、定位、测距,并能通过单片机控制电机实现实时追踪。对乒乓球颜色特征提取和乒乓球中心定位算法进行了研究,采用Lab特征模型对乒乓球颜色进行识别,引入质心法对乒乓球中心实现定位,推理出单摄像头对乒乓球进行距离测量的示意图和理论公式。依据定位中心和距离值,通过单片机控制电机运动,实现对乒乓球追踪功能。经实验测试,该机器人能够对乒乓球进行快速精准定位,算法处理时间小于100ms,定位误差小于30mm。     

一种手机平台控制的轮式探测机器人的设计与实现

本项目建立了一个远程实时控制系统,机器人利用摄像头将其采集到的视频信号压缩后,经过Wi-Fi模块通过无线网络传输技术将数据传送到控制端(手机),解压缩以后反应在控制端平台上,控制者通过这些信息发出相应的控制信号,系统再通过Wi-Fi模块将控制端的控制信号传输到远程终端(机器人),远程终端在ARM平台下的Linux嵌入式系统中对信号进行解析并进行相应操作响应,实现控制端对与轮式探测机器人的实时操控。     

基于Kinect手势识别的应用与研究

为解决当前智能家居系统操作繁琐的问题,同时为获得更简单的控制方式,并增加用户的体验感受,研究了基于Kinect骨骼信息的手势识别技术,并将其融入至智能家居的人机交互系统中。在该系统中,用户可以自定义手势动作或语音实现家居设备的智能控制。使用了一种基于加权动态时间规整的模板匹配手势识别算法。通过Kinect的深度摄像头获取手势深度图像和骨骼图像数据,并采用加权动态时间规整算法进行识别。实验表明使用该算法实现手势识别是可行且有效的,且其最佳识别位置是在Kinect的正前方2～2.5m处,识别准确率达到96%左右。     

基于MSP430F5529的简易风洞控制系统设计

基于MSP430F5529设计了一个简易风洞控制系统。选用超声波传感器检测乒乓球在简易风洞中的位置,微控制器MSP430F5529通过对乒乓球当前位置的分析判断控制直流电机的转速,从而完成对乒乓球升降高度的实时控制。     

一种智能搬运机器人的设计与实现

为了提高工业生产中重复性工作的劳动效率,设计了一种基于ATmega16单片机的智能搬运机器人;系统采用涡流传感器检测金属线进行路径识别,通过单片机控制舵机使机器人能够按照预定的路线行驶,自主判断货物方位,并将货物运送到指定位置:主要阐述了系统中机械结构设计、路径识别、舵机驱动等重要环节的实现方法,并通过ISP下载模块实现系统与电脑的信息交互;通过模型测试表明,该系统能够满足自动搬运货物的设计目的,对降低工人从事重复性劳动时的事故率和提高劳动生产率具有较高的应用价值.     

基于LabVIEW的乒乓球机器人设计

乒乓球机器人系统对实时性的要求非常高,其设计涉及到了运动物体的识别、三维定位、快速轨迹预测和机械臂精确控制.在LabVIEW中调用图像处理模块获取并处理图像,满足实时性要求;使用滑轨和机械手结合来去球的方案,扩大了击球范围;调用Modbus函数库简化了伺服电机驱动器的控制,制作的样机目前已经可以实现乒乓球三维定位、击球点预测以及几个回合的人机对打.     

基于红外小目标识别的空中鼠标解决方案

该文提出一种基于运动物体DBT(detect before track)背景消除法识别红外小目标,并将其应用于空中鼠标中。实验主要经过色彩空间转换、阈值分割、图像开运算、连通分析后进行目标识别,并通过摄像头标定获取识别目标在视野中的相对位置。通过单片机以及2.4G通信模块,将除位置信息以外的控制信息传递到电脑。以电脑进行位置运算,单片机进行控制。实验结果表明,该方案识别的目标精度较高,价格低廉,具有一定应用前景。     

面向网络的空间信息提取系统研究

随着网络技术的不断发展,互联网已经成为一个海量、复杂多样的数据源,特别是随着Web2. 0与社交网络的兴起,每个网民都可视为一个空间传感器,其源源不断地将周围的空间信息发布在网上,互联网中的空间信息日益丰富。提出了面向网络的空间信息提取系统,在从Web页面中所包含的半结构文本或自由文本中识别出完整位置的基础上,提取出与该位置相关的专题属性信息,并将其结构化和空间化。通过系统实例的研究,验证了本系统的可行性。     

Application of frequency-following servocompensator to tracking control

A servocompensator with performance of frequency following is applied to a tracking control system whose reference signal includes sinusoidal components. The applied servocompensator is characterized by using signals synchronized with the reference signal in calculating the control input by means of convolution integral. Since the transmission zeros of the closed-loop system are generated at the same locations as the unstable poles of the reference signal automatically, the pole-zero cancellation necessary for asymptotic regulation is achieved. To demonstrate the efficacy of the servocompensator, an experimental apparatus is developed in which the position of a pick up is controlled to hold the distance between the pick up and a rotating disk. In such systems, the misalignment of the disk causes the reference signal to include sinusoidal components. The control error was remarkably reduced at any rotational speed by the servocompensator.     

体育视频中场地自动检测方法

主要针对虚拟广告系统设计了简单而有效的算法来实现体育视频中场地检测,首先通过基于颜色空间直方图统计的方法实现了场地主区域的自动提取,再通过Top-Hat变换和改进的最大类间方差法实现了场地边缘检测,最后通过Hough直线检测和最小二乘拟合相结合的方法精确检测到所有的场地线。大量的实验表明,本文提出的场地检测方法对于羽毛球场地,网球场地,乒乓球场地等都有很好的检测效果,提取结果可用于摄像机定标和比赛场地的重建。     

一种有效的视频数据提取方法

一般的基于FPGA的视频采集系统中,FPGA与视频解码芯片的连接需要用到较多的FPGA I/O资源。文章介绍了一种新的通过检测SAV信号对视频信号进行定位和采集的方法。该方法只需用到视频数据和像素时钟同步信号即可实现有效的视频提取功能,节省了有限的FPGA I/O资源,进一步提高了系统的集成度。仿真结果表明,该方法达到了实时提取视频信号的目的。     

基于ESP32的乒乓球实时运动模式识别

研究一种基于ESP32传感器的乒乓球动作实时识别系统, 适用于乒乓球爱好者在进行乒乓球运动时的动作识别. 该系统能够采集用户运动时的三轴加速度信号并实时传输至计算机端. 在计算机端对原始加速度信号进行数据预处理, 包括滤波降噪和加窗分割, 提取预处理后加速度信号的时域特征, 利用PCA算法实现特征降维, 最后用决策树(decision tree)算法构建最优的学习模型实现对运动的分类和显示. 实验结果表明: 该系统可实时显示乒乓球4种动作的识别与计数, 识别准确率达97.32%.     

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.     

Training a Vision Guided Mobile Robot

This paper presents the design, implementation and evaluation of a trainable vision guided mobile robot. The robot, CORGI, has a CCD camera as its only sensor which it is trained to use for a variety of tasks. The techniques used for training and the choice of natural light vision as the primary sensor makes the methodology immediately applicable to tasks such as trash collection or fruit picking. For example, the robot is readily trained to perform a ball finding task which involves avoiding obstacles and aligning with tennis balls. The robot is able to move at speeds up to 0.8 ms^-1 while performing this task, and has never had a collision in the trained environment. It can process video and update the actuators at 11 Hz using a single $20 microprocessor to perform all computation. Further results are shown to evaluate the system for generalization across unseen domains, fault tolerance and dynamic environments.     

Training a Vision Guided Mobile Robot

This paper presents the design, implementation and evaluation of a trainable vision guided mobile robot. The robot, CORGI, has a CCD camera as its only sensor which it is trained to use for a variety of tasks. The techniques used for train ing and the choice of natural light vision as the primary sensor makes the methodology immediately applicable to tasks such as trash collection or fruit picking. For example, the robot is readily trained to perform a ball finding task which involves avoiding obstacles and aligning with tennis balls. The robot is able to move at speeds up to 0.8 ms^-1 while performing this task, and has never had a collision in the trained environment. It can process video and update the actuators at 11 Hz using a single $20 microprocessor to perform all computation. Further results are shown to evaluate the system for generalization across unseen domains, fault tolerance and dynamic environments.     

Pervasive computing in sports training

Our system involves a table tennis practice table where one-half of the table in the vertical position returns the player's shots. We use a digital projector calibrated with the table's vertical half, a digital video camera, and a vibration detector mounted on the table. When a ball strikes the wall, the vibration detector triggers the camera to photograph the wall. Our system processes the photograph and projects an image of the ball's impact location onto the wall. The system can also project statistics during play. We designed it to be low cost, to minimize setup time, and to maximize portability. We envisage that the system could work for many interactive games and sports, such as regular tennis, darts, and soccer.     

Artificial neural networks for coordination and control: The portability of experiential representations

It is time to locate connectionist representation theory in the new wave of robotics research. The utility of representations developed in artificial neural networks (ANNs) during learning has been demonstrated in cognitive science research since the 1980s. The research reported here puts learned representations to work in a decentered control task, the disembodied arm problem, in which a mobile robot operates an arm fixed to a table to pick up objects. There is no physical linkage between the arm and the robot and so the robot's point of view must be decentered. This is done by developing a modular Artificial Neural Net system in three stages: (i) a classifier net is trained with laser scan data to output transformationally invariant position classes; (ii) an arm net is trained for picking up objects; (iii) an inter net is trained to communicate and coordinate the sensing and acting. The completed system is shown to create new nonsymbolic transformationally invariant representations in order to perform the effective generalization of decentered viewpoints.     

一种实现区域覆盖的图形拾取算法

为了准确拾取区域覆盖情况下的图形，逐步分析并给出了决定图形拾取与否的三个条件：拾取点是否落在图形的可拾取区域内、拾取点到图形的最短距离是否为最小、拾取点所在的图形的拾取区域面积是否最小。通过对这三个条件的依次判断，提出一种将最短距离与面积结合的方式来实现区域覆盖情况下的图形拾取算法，高效而快捷地实现了图形的拾取。