Computer vision techniques for construction safety and health monitoring

For construction safety and health, continuous monitoring of unsafe conditions and action is essential in order to eliminate potential hazards in a timely manner. As a robust and automated means of field observation, computer vision techniques have been applied for the extraction of safety related information from site images and videos, and regarded as effective solutions complementary to current time-consuming and unreliable manual observational practices. Although some research efforts have been directed toward computer vision-based safety and health monitoring, its application in real practice remains premature due to a number of technical issues and research challenges in terms of reliability, accuracy, and applicability. This paper thus reviews previous attempts in construction applications from both technical and practical perspectives in order to understand the current status of computer vision techniques, which in turn suggests the direction of future research in the field of computer vision-based safety and health monitoring. Specifically, this paper categorizes previous studies into three groups—object detection, object tracking, and action recognition—based on types of information required to evaluate unsafe conditions and acts. The results demonstrate that major research challenges include comprehensive scene understanding, varying tracking accuracy by camera position, and action recognition of multiple equipment and workers. In addition, we identified several practical issues including a lack of task-specific and quantifiable metrics to evaluate the extracted information in safety context, technical obstacles due to dynamic conditions at construction sites and privacy issues. These challenges indicate a need for further research in these areas. Accordingly, this paper provides researchers insights into advancing knowledge and techniques for computer vision-based safety and health monitoring, and offers fresh opportunities and considerations to practitioners in understanding and adopting the techniques.     

大尺寸工件视觉测量中的图像拼接方法

为提高大尺寸工件机器视觉测量精度,提出一种基于坐标变换的图像拼接方法,并给出该坐标变换方法的解算模型及求解方法,分析了局部测量误差、重叠区域公共点个数等对坐标变换精度的影响,得到了最佳拼接条件,并通过实验对其进行了验证;实验结果表明,该方法简单易行、结果可靠,对存在平移、旋转以及尺度缩放的图像都具有良好的效果,能有效地解决采用机器视觉进行大尺寸工件精确测量时的坐标归一化问题。     

基于OPENCV的手势识别系统的设计与实现

人与计算机的交互技术是一种新型的计算机技术,且逐渐演变为一种主流技术和计算机领域的技术热点;为了能够更好地识别手势和跟踪手势的运动轨迹,提出了基于OPENCV的手势识别系统,系统引入了OPENCV计算机视觉库,OPENCV作为优秀的计算机视觉库,为设计的实现提供了便捷的代码,利用OPENCV技术中的图像处理算法,首现通过摄像头采集数据图像,并对采集到的图像进行一系列的缩放,去噪以及锐化等处理,然后对人体手势建立肤色模型,然后经过灰度阈值化来转换成二值图像,得到手轮廓的数据图像后,采用轮廓匹配方法识别出手型;最后通过10种基本的手势模型对比验证了本系统具有一定的实时性,并且识别率可以达到95％以上.     

基于激光和单目相机信息融合的智能轮椅避障策略研究

针对智能轮椅使用环境复杂多变,障碍物形状各异,单一传感器无法获得完整的环境信息的问题,提出一种基于激光传感器和单目视觉传感器信息融合的障碍物检测方法。通过单目相机和激光雷达传感器感知智能轮椅周围环境,得到障碍物的形状、距离分布状况等信息;在此基础上提出两种传感器信息的融合策略,建立局部障碍物地图,进一步采用模糊神经网络完成整体避障算法,实现智能轮椅安全、快速避障等功能。实验结果验证了文中所提避障算法的可行性及有效性。     

数码迷彩空间混色的影响因素研究*

针对数码迷彩的空间混色问题开展了一系列的研究实验,分析了各种影响因素与混色的关系。从人眼视觉、空间混色的形成过程入手,通过三原色的混色实验和数码迷彩的混色实验,对其影响因素如色块大小、观测距离等进行研究和量化分析,获得了它们对空间混色影响的基本规律。实验结果对于数码迷彩的设计和优化将提供可靠的参考依据。     

Remote-Controlled Ambidextrous Robot Hand Concept

In the development of robotic limbs, the side of members is of importance to define the shape of artificial limbs and the range of movements. It is mainly significant tbr biomedical applications concerning patients suffering arms or legs injuries, fn this paper, the concept of an ambidextrous design lbr robot hands is introduced. The fingers can curl in one xvay or another, to imitate either a right hand or a left hand. The advantages and inconveniences of different models have been investigated to optimise the range and the maximum force applied by fingers. Besides, a remote control interthce is integrated to the system, allowing both to send comrnands through internet and to display a video streaming of the ambidextrous hand as feedback. Therefore, a robotic prosthesis could be used for the first time in telerehabilitation. The main application areas targeted are physiotherapy alter strokes or management of phantom pains/br amputees by/earning to control the ambidextrous hand. A client application is also accessible on Facehook social network, making the robotic limb easily reachable for the patients. Additionally the ambidextrous hand can be used tbr robotics research as well as artistic performances.     

Virtual reflected-light microscopy

Research on better methods to digitally represent microscopic specimens has increased over recent decades. Opaque specimens, such as microfossils and metallurgic specimens, are often viewed using reflected light microscopy. Existing 3D surface estimation techniques for reflected light microscopy do not model reflectance, restricting the representation to only one illumination condition and making them an imperfect recreation of the experience of using an actual microscope. This paper introduces a virtual reflected-light microscopy (VRLM) system that estimates both shape and reflectance from a set of specimen images. When coupled with anaglyph creation, the system can depict both depth information and illumination cues under any desired lighting configuration. Digital representations are compact and easily viewed in an online setting. A prototype used to construct VRLM representations is comprised only of a microscope, a digital camera, a motorized stage and software. Such a system automatically acquires VRLM representations of large batches of specimens. VRLM representations are then disseminated in an interactive online environment, which allows users to change the virtual light source direction and type. Experiments demonstrate high quality VRLM representations of 500 microfossils.     

Ergonomic Analysis of Head‐Mounted Night Vision Goggle Systems in Simulated Ground Operations

This study compared the mounting time and strain on the head‐neck while using an old model (forehead and cheek supported) and a new model (head harness supported) of head‐mounted night vision goggle systems in the postures of standing, sitting, and lying prone. The results showed that effectively reducing the mounting time was attributed to the ease and less time required for length adjustments of the head harness in the new model. The novel device received higher positive subjective ratings for convenience and comfort while wearing than did the old model. Although the weight of the new model was decreased to 0.794 kg and the length was reduced to 0.155 m, the loads on the head‐neck complex remained high. These findings suggest that a rifle should be fitted with the night vision goggle or handheld model to decrease neck‐shoulder workload if the prone position is the most frequent and important posture for soldiers. © 2011 Wiley Periodicals, Inc.     

基于双目视觉的机器人目标定位与机械臂控制

为了更好地与复杂多变的非结构化环境进行交互,完成对目标物体的识别和抓取,提出了一种应用于服务机器人平台的基于双目视觉的仿人机械臂控制方法.文中首先用D-H方法对机械臂进行建模,并对这个模型做了改进,给出了一种更加简便的3+1自由度仿人机械臂的逆解算法,采用基于双目视觉与颜色分割的目标识别方法;然后根据识别出的目标三维坐标信息控制机械臂完成抓取任务;最后,本方法在家庭服务机器人上得到了验证,机器人能够完成对目标物体的识别和抓取动作.     

新型多自由度假手的研制

介绍了一种新型多自由度假手的机械结构、运动方式及其静力学分析．假手的主要结构部件由光敏树 脂制成,使用小型直流伺服电机作为驱动器．手指采用欠驱动及腱驱动方式．该假手具有拟人的外观、较低的整体 重量和对被抓物体外形的适应能力．文章研究了假手的运动特点及接触力规律,通过仿真对假手的接触受力规律进 行了校验．最后通过实验,对假手的运动规律与抓握能力进行了验证．