Measuring elemental time and duty cycle using automated video processing

A marker-less 2D video algorithm measured hand kinematics (location, velocity and acceleration) in a paced repetitive laboratory task for varying hand activity levels (HAL). The decision tree (DT) algorithm identified the trajectory of the hand using spatiotemporal relationships during the exertion and rest states. The feature vector training (FVT) method utilised the k-nearest neighbourhood classifier, trained using a set of samples or the first cycle. The average duty cycle (DC) error using the DT algorithm was 2.7%. The FVT algorithm had an average 3.3% error when trained using the first cycle sample of each repetitive task, and had a 2.8% average error when trained using several representative repetitive cycles. Error for HAL was 0.1 for both algorithms, which was considered negligible. Elemental time, stratified by task and subject, were not statistically different from ground truth (p < 0.05). Both algorithms performed well for automatically measuring elapsed time, DC and HAL.

Practitioner Summary: A completely automated approach for measuring elapsed time and DC was developed using marker-less video tracking and the tracked kinematic record. Such an approach is automatic, repeatable, objective and unobtrusive, and is suitable for evaluating repetitive exertions, muscle fatigue and manual tasks.     

自然环境视频中基于显著鲁棒轨迹的行为识别

目的人类行为识别是计算机视觉领域的一个重要研究课题。由于背景复杂、摄像机抖动等原因,在自然环境视频中识别人类行为存在困难。针对上述问题,提出一种基于显著鲁棒轨迹的人类行为识别算法。方法该算法使用稠密光流技术在多尺度空间中跟踪显著特征点,并使用梯度直方图(HOG)、光流直方图(HOF)和运动边界直方图(MBH)特征描述显著轨迹。为了有效消除摄像机运动带来的影响,使用基于自适应背景分割的摄像机运动估计技术增强显著轨迹的鲁棒性。然后,对于每一类特征分别使用Fisher Vector模型将一个视频表示为一个Fisher向量,并使用线性支持向量机对视频进行分类。结果在4个公开数据集上,显著轨迹算法比Dense轨迹算法的实验结果平均高1%。增加摄像机运动消除技术后,显著鲁棒轨迹算法比显著轨迹算法的实验结果平均高2%。在4个数据集(即Hollywood2、You Tube、Olympic Sports和UCF50)上,显著鲁棒轨迹算法的实验结果分别是65.8%、91.6%、93.6%和92.1%,比目前最好的实验结果分别高1.5%、2.6%、2.5%和0.9%。结论实验结果表明,该算法能够有效地识别自然环境视频中的人类行为,并且具有较低的时间复杂度。     

A comparison of risk assessment of single and combination manual handling tasks: 2. Discomfort,Rating of Perceived Exertion and heart rate measures

Although many manual handling activities involve combinations of pull, lift, carry, lower and push, there are few reports of investigation of how to assess the risk in these combination tasks. Two strategies have been suggested in the literature for estimating the risk in a combination task based on the measures of the separate components of that task. The aim of the study was to compare the risks assessed in single manual handling tasks with those in combination tasks. Ratings of discomfort, exertion and heart rate were collected from nine male and nine female students, performing combination and single tasks. Combination tasks consisted of sequences of pull, lift, carry, lower and push tasks. Combination tasks were performed at 1.min^?1 and 3.min^?1 whilst single tasks (lift, lower, push, pull and carry) were performed at 3.min^?1 and 6.min^?1. The rating of exertion and heart rate for each combination task was compared to the exertion rating and heart rate of the single tasks which comprised the combination task using repeated measures anaylsis of variance with specified contrasts. Similar comparisons for the discomfort data were performed using Friedman and Wilcoxon tests. In at least one of the twelve comparisons performed for each dependent variable, the combination task value was significantly different to each single task value. The differences occurred regardless of whether the most critical single task value or an average of all single task values was used. It was concluded that the risk in combination manual tasks can not be accurately assessed by using estimates from discomfort, exertion ratings and heart rate measures of single tasks.     

The accuracy of conventional 2D video for quantifying upper limb kinematics in repetitive motion occupational tasks

Marker-less 2D video tracking was studied as a practical means to measure upper limb kinematics for ergonomics evaluations. Hand activity level (HAL) can be estimated from speed and duty cycle. Accuracy was measured using a cross-correlation template-matching algorithm for tracking a region of interest on the upper extremities. Ten participants performed a paced load transfer task while varying HAL (2, 4, and 5) and load (2.2 N, 8.9 N and 17.8 N). Speed and acceleration measured from 2D video were compared against ground truth measurements using 3D infrared motion capture. The median absolute difference between 2D video and 3D motion capture was 86.5 mm/s for speed, and 591 mm/s² for acceleration, and less than 93 mm/s for speed and 656 mm/s² for acceleration when camera pan and tilt were within ± 30 degrees. Single-camera 2D video had sufficient accuracy ( < 100 mm/s) for evaluating HAL.

Practitioner Summary: This study demonstrated that 2D video tracking had sufficient accuracy to measure HAL for ascertaining the American Conference of Government Industrial Hygienists Threshold Limit Value^® for repetitive motion when the camera is located within ± 30 degrees off the plane of motion when compared against 3D motion capture for a simulated repetitive motion task.     

Relationships between observational estimates and physical measurements of upper limb activity †

This study examined the internal validity of observational-based ergonomic job analysis methods for assessing upper limb force exertion and repetitive motion. Six manual tasks were performed by multiple ‘workers’ while direct measurements were made to quantify force exertion and kinematics of the upper limb. Observational-based analyses of force and upper limb motion/repetition were conducted by 29 professional ergonomists. These analysts overestimated the magnitude of individual force exertions – temporal aspects of force exertion (duty cycle) were estimated more accurately. Estimates of the relative severity of repetitive motions among the jobs were accurate. Absolute counts of repetitive motions were less accurate. Modest correlations (r² = 0.28 to r² = 0.50) were observed between ratings of hand activity level and measured joint velocities. Ergonomic job analyses relying on systematic observation should be applied and interpreted with consideration given to the capabilities and limitations of analysts in estimating the physical risk factors. These findings are relevant to a better understanding of the internal validity of ergonomic job analysis methods based on systematic observation.     

Biomechanical analysis of upper trapezius,erector spinae and brachioradialis fatigue in repetitive manual packaging tasks: Evidence from Chinese express industry workers

This study aims to assess the effects of repetitive motion-induced fatigue during manual packaging on kinematics and muscle activity of the arm, back and shoulder. Eighteen participants performed a 60-min manual packaging task. Electromyography median frequency of RUT, RES, LES and RB decreased by 13.2% (p < 0.05), 12.8% (p < 0.05), 11.3% (p < 0.05) and 21% (p < 0.001), respectively over time. The head flexion and the right upper arm flexion angles significantly (p < 0.001) differed among different packaging sizes, and similar results were observed for left upper arm flexion and rotation. The trunk and pelvis angles significantly (p < 0.05) differed over time under three kinds of packaging. The results showed there was an increase in the rating of perceived exertion from 8.56 to 16.94 (p<0.01) during the task. The outcomes of this study indicate repetitive movements in the manual packaging task resulted in elevated prevalence of muscle discomfort in packaging workers, especially the right brachioradialis. Interventions such as specific work: rest ratios, workplace redesign should be explored to relieve muscle fatigue and discomfort.Relevance to industryFor workers performing repetitive manual packaging tasks, biomechanical analysis of different muscles groups can help in developing appropriate ergonomic interventions.     

The Revised Strain Index: an improved upper extremity exposure assessment model

Abstract

The Revised Strain Index (RSI) is a distal upper extremity (DUE) physical exposure assessment model based on: intensity of exertion, frequency of exertion, duration per exertion, hand/wrist posture and duration of task per day. The RSI improves upon the 1995 Strain Index (SI) by using continuous rather than categorical multipliers, and replacing duty cycle with duration per exertion. In a simulation of 13,944 tasks, the RSI and 1995 SI showed good agreement in risk predictions for 1995 SI scores ≤3 (safe) and >13.5 (hazardous). For tasks with 1995 SI scores of >3 and ≤13.5, the two models showed marked disagreement, with the RSI providing much greater discriminations between ‘safe’ and ‘hazardous’ tasks for various combinations of force, repetition and duty cycle. We believe the RSI is a substantially improved model that will be useful for DUE task analysis, intervention and design.

Practitioner Summary: RSI is a substantial improvement over the 1995 SI. It should be a valuable tool for designing and analysing tasks to determine risk of musculoskeletal injuries. RSI is applicable to a wide variety of tasks including very low force and very high repetition tasks such as keyboard use.     

The effect of bracing availability on one-hand isometric force exertion capability

Environmental obstructions that workers encounter can kinematically limit the postures that they can achieve. However, such obstructions can also provide an opportunity for additional support by bracing with the hand, thigh or other body part. The reaction forces on bracing surfaces, which are in addition to those acting at the feet and task hand, are hypothesised to improve force exertion capability, and become required inputs to biomechanical analysis of tasks with bracing. The effects of kinematic constraints and associated bracing opportunities on isometric hand force were quantified in a laboratory study of 22 men and women. Analyses of one-hand maximal push, pull and lift tasks demonstrated that bracing surfaces available at the thighs and non-task hand enabled participants to exert an average of 43% more force at the task hand. Task hand force direction deviated significantly from the nominal direction for exertions performed with bracing at both medium and low task hand locations.

Practitioner summary: This study quantifies the effect of bracing on kinematically constrained force exertions. Knowledge that appropriate bracing surfaces can substantially increase hand force is critical to the evaluation of task-oriented strength capability. Force estimates may also involve large off-axis components, which have clear implications for ergonomic analyses of manual tasks.     

光场成像技术及其在计算机视觉中的应用

目的 光场成像技术刚刚在计算机视觉研究中展开初步应用,其相关研究比较零散,缺乏系统性。本文旨在系统介绍光场成像技术发展以及其应用在计算机视觉研究中有代表性的工作。方法 从解决计算机视觉问题的角度出发,4个层面讨论光场成像技术最近十年的研究工作,包括:1)主流的光场成像设备及其作为计算机视觉传感器的优点与不足;2)光场相机作为视觉传感器的标定、解码以及预处理方法;3)基于4维光场的图像渲染与重建技术,以及其如何促进计算机视觉研究;4)以4维光场数据为基础的特征表达方法。结果 逐层梳理出光场成像在求解视觉问题中的优势和局限,分析其中根本性的原理与掣肘,力图总结出亟待解决的关键问题以及未来的发展趋势。结论 作为一种颇具前景的新型计算机视觉传感器技术,光场成像技术的研究必将更为广泛和深入。研究应用于计算机视觉的光场成像技术将有力的引导和促进计算机视觉和光场成像技术协同发展。     

Temporal video segmentation by event detection: A novelty detection approach

Temporal segmentation of videos into meaningful image sequences containing some particular activities is an interesting problem in computer vision. We present a novel algorithm to achieve this semantic video segmentation. The segmentation task is accomplished through event detection in a frame-by-frame processing setup. We propose using one-class classification (OCC) techniques to detect events that indicate a new segment, since they have been proved to be successful in object classification and they allow for unsupervised event detection in a natural way. Various OCC schemes have been tested and compared, and additionally, an approach based on the temporal self-similarity maps (TSSMs) is also presented. The testing was done on a challenging publicly available thermal video dataset. The results are promising and show the suitability of our approaches for the task of temporal video segmentation.     

Effect of a novel teat preparation system on upper extremity muscle activity among U.S. large-herd dairy parlor workers

BackgroundDairy production in the U.S. is moving towards large-herd milking operations resulting in an increase in task specialization and work demands. Large-herd dairy parlor workers experience a high prevalence of musculoskeletal symptoms in the upper extremity. The purpose of this study is to evaluate the effects of an alternative teat scrubber (TS) cow preparation method on upper extremity muscle activity among large-herd parlor workers, as well compare to muscle activity associated with conventional manual milking tasks.MethodsUpper extremity muscle activity was recorded among U.S. large-herd parlor workers (n = 15) using surface electromyography. Participants performed multiple task cycles, using both conventional and TS methods. Muscle activity levels were then compared across conventional manual and TS milking tasks.ResultsConventional manual milking tasks of dip, strip and wipe were associated higher muscle activity levels of the upper trapezius and anterior deltoid. Biceps muscle activity was greatest during teat dip and wipe. Forearm flexor and extensor muscle activity was greatest during teat wipe and dip. The TS system resulted in more desirable anterior deltoid EMG profiles, and less desirable biceps, forearm flexor and extensor profiles.ConclusionsResults suggest that the TS system is effective in reducing anterior deltoid muscle activation levels. The TS system also appears to result in increased biceps, forearm flexor and extensor muscle activation levels. Increases in muscle activation levels could be offset by reduced repetitiveness resulting from three conventional manual milking tasks being replaced with one TS task.Relevance to industryIf parlor production requirements (e.g., quality and onsistency) are maintained while simultaneously reducing cumulative muscle loading and worker fatigue, then TS use should be considered in milking parlor operations.     

Quantifying the physical intensity of construction workers,a mechanical energy approach

Construction workers typically undertake highly demanding physical tasks involving various types of stresses from awkward postures, using excessive force, highly repetitive actions, and excessive energy expenditure, which increases the likelihood of unsafe actions, productivity loss, and human errors. Biomechanical models have been developed to estimate joint loadings, which can help avoid strenuous physical exertion, potentially enhancing construction workforce productivity, safety, and well-being. However, the models used are mainly in 2D, or to predict static strength ignored their velocity and acceleration or using marker-based method for dynamic motion data collection. To address this issue, this paper proposes a novel framework for investigating the mechanical energy expenditure (MEE) of workers using a 3D biomechanical model based on computer vision-based techniques. Human 3D Pose Estimation algorithm based on 2D videos is applied to approximate the coordinates of human joints for working postures, and smart insoles are used to collect foot pressures and plantar accelerations, as input data for the biomechanical analyses. The results show a detailed MEE rate for the whole body, at which joints the maximum and minimum values were obtained to avoid excessive physical exertion. The proposed method can approximate the total daily MEE of construction tasks by summing the assumed cost of individual tasks (such as walking, lifting, and stooping), providing suggestions for the design of a daily workload that workers can sustain without developing cumulative fatigue.     

Automatically propagating changes from reference implementations to code generation templates

ContextCode generators can automatically perform some tedious and error-prone implementation tasks, increasing productivity and quality in the software development process. Most code generators are based on templates, which are fundamentally composed of text expansion statements. To build templates, the code of an existing, tested and validated implementation may serve as reference, in a process known as templatization. With the dynamics of software evolution/maintenance and the need for performing changes in the code generation templates, there is a loss of synchronism between the templates and this reference code. Additional effort is required to keep them synchronized.ObjectiveThis paper proposes automation as a way to reduce the extra effort needed to keep templates and reference code synchronized.MethodA mechanism was developed to semi-automatically detect and propagate changes from reference code to templates, keeping them synchronized with less effort. The mechanism was also submitted to an empirical evaluation to analyze its effects in terms of effort reduction during maintenance/evolution templatization.ResultsIt was observed that the developed mechanism can lead to a 50% reduction in the effort needed to perform maintenance/evolution templatization, when compared to a manual approach. It was also observed that this effect depends on the nature of the evolution/maintenance task, since for one of the tasks there was no observable advantage in using the mechanism. However, further studies are needed to better characterize these tasks.ConclusionAlthough there is still room for improvement, the results indicate that automation can be used to reduce effort and cost in the maintenance and evolution of a template-based code generation infrastructure.     

Mental workload and cognitive task automaticity: an evaluation of subjective and time estimation metrics

Abstract

The evaluation of mental workload is becoming increasingly important in system design and analysis. The present study examined the structure and assessment of mental workload in performing decision and monitoring tasks by focusing on two mental workload measurements: subjective assessment and time estimation. The task required the assignment of a series of incoming customers to the shortest of three parallel service lines displayed on a computer monitor. The subject was either in charge of the customer assignment (manual mode) or was monitoring an automated system performing the same task (automatic mode). In both cases, the subjects were required to detect the non-optimal assignments that they or the computer had made. Time pressure was manipulated by the experimenter to create fast and slow conditions. The results revealed a multi-dimensional structure of mental workload and a multi-step process of subjective workload assessment. The results also indicated that subjective workload was more influenced by the subject's participatory mode than by the factor of task speed. The time estimation intervals produced while performing the decision and monitoring tasks had significantly greater length and larger variability than those produced while either performing no other tasks or performing a well practised customer assignment task. This result seemed to indicate that time estimation was sensitive to the presence of perceptual/cognitive demands, but not to response related activities to which behavioural automaticity has developed.     

Effect of physical workload and modality of information presentation on pattern recognition and navigation task performance by high-fit young males

Abstract

Many occupations require both physical exertion and cognitive task performance. Knowledge of any interaction between physical demands and modalities of cognitive task information presentation can provide a basis for optimising performance. This study examined the effect of physical exertion and modality of information presentation on pattern recognition and navigation-related information processing. Results indicated males of equivalent high fitness, between the ages of 18 and 34, rely more on visual cues vs auditory or haptic for pattern recognition when exertion level is high. We found that navigation response time was shorter under low and medium exertion levels as compared to high intensity. Navigation accuracy was lower under high level exertion compared to medium and low levels. In general, findings indicated that use of the haptic modality for cognitive task cueing decreased accuracy in pattern recognition responses.

Practitioner Summary: An examination was conducted on the effect of physical exertion and information presentation modality in pattern recognition and navigation. In occupations requiring information presentation to workers, who are simultaneously performing a physical task, the visual modality appears most effective under high level exertion while haptic cueing degrades performance.     

照片水印自动排版与配色

目的 随着用户对图像传播的个性化要求不断提升,图像-与文本融合成为了目前传播的新形式。为用户照片自动地添加具有美观的水印文字成为图像处理的一个新方向-视觉媒体的自动设计。通过结合设计学原理和计算机视觉原理,为水印文字块在图像中的位置和配色进行优化。方法 基于视觉显著性理论及构图学理论,为水印文字块确定最佳位置。然后根据视觉反差和颜色和谐化理论为文字选取合适颜色。为更完美呈现文字与图像的和谐关系,基于色彩和谐化理论对原图像进行自适应的色轮模板匹配,并以此选取文本色彩。结果 为验证本文所提出的算法,采用用户调查方式分别与腾讯的水印相机及华夏地理（国家地理杂志中文网）配图作对比。与腾讯的水印相机相比,排版与配色后的结果优于仅和谐化后结果（73.25%：17.42%）,同时也优于水印相机原图（73.25%：9.32%）。与华夏地理编辑人工处理后的图片相比,本算法远胜于原始推送结果（97.2%：2.8%）。本算法可以全自动、快捷为用户提供满意的输出结果。结论 无论在与目前自动化的水印相机相比还是简单的人工编辑排版相比,本文算法在提高图文混合排版的设计感及美感上有一定的作用。     

Manual patient transfers: factors that influence decisions and kinematic strategies employed by nursing aides

While extensive literature has characterised factors that influence the acceptable mass of ‘boxes’ during MMH tasks, less is known about these factors when moving ‘people’ in healthcare settings. This study examined factors that influence decisions/approaches employed during manual patient transfers. Sixteen nursing aides manually-transferred a standardised ‘patient’; patient mass was adjusted (using a weight vest) to determine a maximum acceptable patient mass for this task (mass_max). Grip strength was the only worker characteristic significantly associated with mass_max (r?=?0.48). Older worker age was associated with smaller peak trunk flexion (r?= ?0.58) and shoulder abduction (r?= ?0.59), and greater trunk axial twist (r?=?0.52). Workers emphasised that patient characteristics (e.g. physical/cognitive status) influenced their decisions when performing transfers. These findings extend previous literature by suggesting that grip strength is a useful predictor of perceived work capacity, older workers adapt protective postural strategies during patient transfers and worker-patient dynamics are crucial during this high-risk occupational task.

Practitioner Summary: This study examined manual patient transfers performed by nursing aides. Worker grip strength (but not age or size) was associated with perceptions of maximum acceptable patient mass. Kinematic changes suggested more conservative strategies used by older workers. Workers emphasised that patient characteristics substantially influenced their decisions when performing transfer tasks.     

Adaptive Video Fast Forward

We derive a statistical graphical model of video scenes with multiple, possibly occluded objects that can be efficiently used for tasks related to video search, browsing and retrieval. The model is trained on query (target) clip selected by the user. Shot retrieval process is based on the likelihood of a video frame under generative model. Instead of using a combination of weighted Euclidean distances as a shot similarity measure, the likelihood model automatically separates and balances various causes of variability in video, including occlusion, appearance change and motion. Thus, we overcome tedious and complex user interventions required in previous studies. We use the model in the adaptive video forward application that adapts video playback speed to the likelihood of the data. The similarity measure of each candidate clip to the target clip defines the playback speed. Given a query, the video is played at a higher speed as long as video content has low likelihood, and when frames similar to the query clip start to come in, the video playback rate drops. Set of experiments o12n typical home videos demonstrate performance, easiness and utility of our application.Nemanja Petrovic received his Ph.D. from University of Illinois in 2004. He is currently the member of research staff at Siemens Corporate Research at Princeton, New Jersey. His professional interests are computer vision and machine learning. Dr. Petrovic has published 20 papers in the area of video understanding, data clustering and enhancement.Nebojsa Jojic received his Ph.D. from University of Illinois at 2001. His currently a researcher at Microsoft Research at Redmond, Washington. His professional interest include computer vision and machine learning. Dr. Jojic has published over 40 papers in the area of computer vision, bioinformatics and graphical models.Thomas Huang received his Sc.D. from MIT in 1963. He is William L. Everitt Distinguished Professor in the University of Illinois, Department of Electrical and Computer Engineering and a full-time faculty member in the Beckman Institute Image Formation and Processing and Artificial Intelligence groups. Professor Huang has published over 600 papers in the area of computer vision, image compression and enhancement, pattern recognition, and multimodal signal processing. He is a Member of the National Academy of Engineering, Foreign Member of the Chinese Academy of Engineering and Chinese Academy of Science, and recipient of IEEE Jack S. Kilby Signal Processing Medal.     

Variation at work: alternations between physically and mentally demanding tasks in blue-collar occupations

Abstract

The aims of this questionnaire study were to describe the occurrence and desired number of alternations between mental and physical tasks in industrial and non-industrial blue-collar work, and determine to which extent selected personal and occupational factors influence these conditions. On average, the 122 participating workers (55 females) reported to have close to four alternations per day between mental and physical tasks, and to desire more alternations than they actually had. They also expressed a general preference for performing a physical task after a mental task and vice versa. In univariate regression models, the desired change in task alternations was significantly associated with gender, age, occupation, years with current work tasks and perceived job control, while occupation was the only significant determinant in a multiple regression model including all factors. Our results suggest that alternations between productive physical and mental tasks could be a viable option in future job rotation.

Practitioner Summary: We addressed attitudes among blue-collar workers to alternations between physically and mentally demanding tasks. More alternations were desired than those occurring in the job, and workers preferred performing a physical task after a mental and vice versa. Alternating physical and mental tasks could, thus, be a viable option in job rotation.