基于创意序列学习的艺术风格学习与绘制系统

基于学习的智能图像风格绘制是当前多媒体特别是艺术风格化领域的一个热门研究课题.图像风格学习算法主要研究利用多媒体数据及机器学习方法对真实样本数据进行自动艺术绘制智能处理.目前主流方法主要针对艺术图像的静态样本进行学习.然而,由于静态数据类中包含的信息是平面化、局部化和非连续化的,难以保证风格化处理的全局一致性.本文旨在针对创意过程中的大规模复杂多媒体数据,从序列任务学习理论上提出一套智能艺术风格绘制的理论模型、设计方法以及优化方法三个层面.本文贡献主要体现在以下研究工作：（1）提出了一套面向数字美术的多媒体数据采集设备与软件系统,（2）利用IRL算法实现了艺术风格行为的模型化及其数字化保护方法,（3）提出基于PGPE的正则化策略学习方法以提高风格学习过程的稳定性.实验结果表明本文提出的方法行之有效地实现针对具体个性风格的照片水墨画艺术风格转化.基于序列化多媒体数据采集与分析,本文提出的面向移动互联网的自动艺术风格绘制辅助系统,不仅具有理论上的创新而且还具有实际上的巨大应用和价值.     

Mushaca: A 3-Degrees-of-Freedom Mouse Supporting Rotation

Based on kinesiology research demonstrating that translation and rotation are inseparable actions in the physical world, we present Mushaca, a 3-degrees-of-freedom mouse that senses rotation in addition to traditional planar position. We present an optical realization of the Mushaca device based on two optical sensors and then evaluate the device through a series of controlled experiments. Our results show that rotation is indeed a useful input modality for a pointing device, and also give some insight into how users perceive the changing coordinate system of the rotating mouse and adapt to this change through kinesthetic learning.     

一种用于图形设备接口的笔画模型方法

视频监控的在屏交互过程中,需要快速地对交互控件、感兴趣区域、活动标记等图形和图像进行绘制,为避免画面的停滞感,要求图形设备接口能快速进行绘制.为此,提出了一种笔画模型方法.利用 DirectX的原始绘制能力,该方法将绘制任务转换成笔画对象,使视频系统在实时运行过程中,只需将已转换的笔画对象的绘制数据提交给 DirectX 进行渲染和显示即可,降低重复绘制时间,满足系统的快速绘制要求.基于笔画模型方法,开发了一种可快速绘制的图形设备接口.     

UCS在AutoCAD三维绘图中的应用和技巧

用户坐标系(UCS)的三维变换是AutoCAD三维绘图的关键,它能为操作人员绘图带来极大方便。但是UCS的灵活多变性往往让初学者很难把握。本文介绍了UCS的定义和使用方法,并通过实例介绍了UCS在实体建模中的具体应用和技巧。     

一种基于位置关系判定的GPU 三维几何图元拾取方法

针对交互式图形应用对拾取在效率、适用范围和拾取信息方面的需求,提出一种 新的基于GPU 的三维几何图元拾取方法。在进行拾取绘制时关闭光栅化,将鼠标位置信息和图 元顶点坐标变换到规范化设备坐标系,通过在几何处理器中判断投影后的二维图元与鼠标或选 择框的位置关系进行命中判定,并利用变换反馈将拾取信息返回应用程序。在提出方法的基础 上,介绍了单体拾取和块拾取的OpenGL 实现。实验表明,该方法的单体拾取效率较基于GPU 的射线相交法有约10%的提升,同时能支持高效块拾取。     

基于IOS移动终端的协同绘图系统

为了增加不同用户绘图的互动性和协作性,基于IOS移动终端通过采用移动CSCW的方法,设计和实现了一个协同绘图系统. 该系统采用了C/S服务模式,通过对移动CSCW的协同控制等技术进行研究,实现了移动终端和服务器之间的信息交互. 同时,结合用户单手绘图的需求,充分利用智能设备中陀螺仪、加速计等传感器的特性,设计了以图元笔画为单元的绘图方案. 在真机上进行了调试,程序运行流畅,达到了设计目标.     

轴测投影的若干问题

重新审视了轴测投影定义的表述。轴测投影法是用平行投影的方法,沿投射方向 将空间坐标系投射到投影面上得到轴测坐标系并进而在这个坐标系下表述空间形体,用这种方 法得到的图纸叫轴测图。三视图用3 个视图分别表示形体的3 个方向(面),轴测图用一个视图 同时表示三直三面角,其本质都是要表达出形体的3 个方向(面),在平面上表述空间形体。讨 论了轴测投影中的一些根本问题,包括平行投影体系、轴测三角形、轴测投影体系、轴测投影 的基本要素和基本公式,剖析了正轴测投影和斜轴测投影之共性和不同,给出了其原理图。     

一种大屏幕人机交互系统的实现方法

针对利用双目立体视觉技术实现大屏幕人机交互系统时摄像头标定复杂的问题, 提出一种大屏幕人机交互系统的实现方法。由两个摄像头负责采集用户手部的图像,获取指尖的二维图像位置坐标,通过透视投影模型,由测量所得的两个摄像头之间的距离、大屏幕左边与摄像头之间的距离、摄像头离地面的高度、大屏幕底边离地面的高度等数据,求出地面上与指尖成像在同一位置的地面坐标,由摄像头的位置及地面坐标求出一条直线,两条直线相交获取指尖空间坐标,由指尖空间坐标选取与屏幕对应的二维平面坐标,经物理坐标到逻辑坐标转换求出手指指向屏幕的实际坐标,实现大屏幕的定位。定位完成后,采用指尖检测方法检测视频图像中指尖是否存在,以此判断食指弯曲和伸开的点击操作。指尖空间坐标的计算方法虽然存在误差,但由于用户在操作过程中能够实时看到鼠标的位置,因此该误差对用户是透明的,而这种方法以简单的、近似的求解避免了复杂的摄像头标定过程,便于系统工程的实施。     

Immersive Whiteboard Collaborative System

An immersive whiteboard system is presented where users at multiple locations can communicate with each other. The system features a virtual environment with vivid avatars, stroke compression and streaming technology to effectively deliver stroke data across meeting participants, friendly human interaction and navigation, virtual and physical whiteboard. The whiteboard is both a physical platform for our input/output interfaces and a virtual screen for sharing common multimedia. It is this whiteboard correspondence that allows the user to physically write on the virtual whiteboard. In addition to drawing on the shared virtual board, the immersive whiteboard in our setup permits users to control the application menus, insert multimedia objects into the world, and navigate around the virtual environment. By integrating multimedia objects and avatar representations into an immersive environment, we provide the users with a more transparent medium so that they feel as if they are communicating and interacting face-to-face. The whiteboard efficiently pulls all the collaboration technologies together. The goal of this collaborative system is to provide a convenient environment for participants to interact with each other and support collaborative applications such as instant messaging, distance learning and conferencing.     

A study on how usability flaws in GUI design increase mouse movements and consequently may affect users' health

The objective of this study is to discuss how software usability flaws may cause a significant increase in mouse movements and, as a potential side effect, may even affect users' health. During the literature review, this article examines the potential relationship between mouse movement and musculoskeletal disorders of the upper extremity, based on studies from medical sciences. Subsequently, in the main part of the present study, three software products were selected that had at least one usability flaw related to extra mouse movement (the selection of the software was made out of over 20 software programs that had been evaluated in our Software Quality Assessment Laboratory). For these products, all additional mouse movements were measured, involving actual users in various settings and computer configurations. The findings showed that even a single usability flaw may increase mouse movement to a magnitude of between 3.6 and 4.7 m/h. The article concludes that the role of software engineering is to focus on software usability as well, taking into account that a user friendly graphical user interface (GUI) which is able to eliminate unnecessary mouse movement may also eventually contribute to the reduction of fatigue and discomfort, caused by musculoskeletal disorders.     

基于单目摄像头的3D动态手势交互

传统的手势交互都需要借助于Leap Motion或Kinect等专用交互设备。以图像通道转换、二值化等图像处理方式提取手势,以手势平面坐标值的变化获取手势的平面移动信息,以手势面积的变化解决了手势深度的问题。通过绘制手势轮廓结合自创的图像匹配算法计算不同图像的匹配率,用最高匹配率选择相对应的手势运动信息。通过摄像头坐标系到3D场景坐标系之间的转换,结合三维图形的几何变换计算变换矩阵,实现手的空间移动与旋转。在不借助专用的手势交互设备的情况下,实现单目摄像头的动态手势交互。     

基于压力传感器与摄像头的颈椎病预防系统

本文设计并实现了一种基于压力传感器与摄像头的颈椎病预防系统.该系统将薄膜压力传感器放置于坐垫中对人体体压进行采集,并使用摄像头采集人类脸部数据.对于坐垫,我们对传感器的布置和数目进行了优化.对于摄像头,我们在多目标识别时通过五官识别减少了系统运算量.在将压力传感器与摄像头两类数据融合之后,我们实现了比单类数据更准确的坐姿识别,进而对颈椎病的危险程度进行估计.我们还使用卷积神经网络实现了对用户的动作检测,可引导用户完成一系列动作用于评估颈椎病危险程度.系统实现小巧美观,采用非侵入式的方法,非常适合于在办公室环境下的颈椎病早期预防.     

Segmentation of lip region in color images by fuzzy clustering

Extraction of the lip region is essential to lip reading, which is a field of image processing that is used to obtain meaningful information by the analysis of lip movement from human face images. Many conventional methods for extraction of the lip region have been proposed. One is to identify the lip position by using geometric face structure. Another discriminates lip and skin regions by using color information only. The former is more complex than the latter; however, it can analyze black and white images as well as color images. The latter is simpler than the former; however, it is difficult to discriminate lip and skin regions because of the high similarity between these two regions, and it is less accurate than the former. Conventional methods usually analyze color coordinate systems to extract lip regions rather than analyzing the coordinate system itself. In this paper, the best color coordinate system for lip extraction was selected by the analysis of discriminability. Segmentation of the lip region with this coordinate system and a new feature vector are proposed.     

Restoration of traditional Chinese shadow play‐Piying art from tangible interaction

Piying, the world's intangible culture heritage, is a characteristic Chinese folk shadow play and one of the origins of modern movie. The spirit of traditional Piying is to express rich emotion and stories through impromptu action change by professional artists. Now Piying gradually fades away in people's lives, encountering the risk of extinction. We focus on transforming the traditional Piying play into an interactive system, which can be seamlessly performed by ordinary people and bring a strong sense of immersion to the users. Two interactive systems were developed, with Kinect‐based interaction or sensors, users can create own digital Piying animations by employing their body movement as input. A field study was presented with 20 students in a primary school. The result showed that our system was far more effective in emotion induction and Piying understanding than traditional one. We demonstrated this system in a charitable foundation and a workshop of a tangible conference. It is also honored to be collected by China's intangible cultural heritage network. Copyright © 2013 John Wiley & Sons, Ltd.     

The Effect of Target Precuing on Pointing With Mouse and Touchpad

In point-and-click interfaces the location of targets is sometimes known to the user before visually identifying it, and sometimes not. This experiment investigates how pointing is affected by whether the target location is precued so that users know it in advance or nonprecued so that users learn it only at the onset of pointing trials. The study investigates this for young, adult, and elderly participants pointing with mouse and touchpad. Target precuing affects the trial completion time, the reaction time, the sheer movement time, and multiple movement kinematics. In addition, target precuing interacts with the use of either mouse or touchpad, with target distance, and with target size, but surprisingly little with participant age. Because the target location was always made known to participants no later than at the onset of the pointing trial, the effects of target precuing must be due to the different possibilities for mental and motor preparations.     

Alternative computer mouse design and testing to reduce finger extensor muscle activity during mouse use

OBJECTIVE: The purpose of this study was to design and test alternative computer mouse designs that attempted to reduce extensor muscle loading of the index and middle fingers by altering the orientation of the button switch direction and the force of the switch. BACKGROUND: Computer users of two-button mouse designs exhibit sustained lifted finger behaviors above the buttons, which may contribute to hand and forearm musculoskeletal pain associated with intensive mouse use. METHODS: In a repeated-measures laboratory experiment, 20 participants completed point-and-click, steering, and drag tasks with four alternative mouse designs and a reference mouse. Intramuscular and surface electromyography (EMG) measured muscle loading, and movement times recorded by software provided a measure of performance. RESULTS: Changing the direction of the switch from a conventional downward to a forward design reduced (up to 2.5% maximum voluntary contraction [MVC]) sustained muscle activity (10th percentile EMG amplitude distribution) in the finger extensors but increased (up to 0.6% MVC) flexor EMG and increased movement times (up to 31%) compared with the reference mouse (p < .001). Implementing a high switch force design also increased flexor EMG but did not differ in movement times compared with the reference mouse (p < .001). CONCLUSION: The alternative mouse designs with altered switch direction reduced sustained extensor muscle loading; however, trade-offs with higher flexor muscle loading and lower performance existed. APPLICATION: Potential applications of this study include ergonomic and human computer interface design strategies in reducing the exposure to risk factors that may lead to upper extremity musculoskeletal disorders.     

霍尔传感器在气动健身器材中的应用

为了能更好地掌握运动者的运动力度,尽可能地减小运动中由于运动过剧造成的损伤,设计了一种气动健身器材,通过实际的气缸压力与霍尔传感器检测出气缸活塞的位移来计算器材使用者所承受的拉力,并结合记录气缸活塞的运动次数来计算运动者消耗的热量,更直观地显示出运动者的运动情况。经过测试,霍尔传感器反应灵敏,检测结果准确,系统稳定性良好,有较高的实用性。     

基于上下文的在线草图识别方法

支持用户自由连贯地绘制草图是在线草图识别研究的目标之一。本文提出一种使用空间关系等上下文和贝叶斯分类器进行草图识别的方法,其主要特点包括两个方面：第一,使用笔画空间关系进行自动笔画成组,可以在不打扰用户绘图的情况下完成候选符号的选取;第二,利用贝叶斯分类器实现在线草图识别,可以解决识别方法的领域相关性及画法敏感性问题。实验验证了方法的有效性。