Quantitative evaluation of stereoscopic augmented reality visualization using an optical see-through head-mounted display

可视化 增强现实 立体视 计算机科学 计算机视觉 光学头戴式显示器 成像体模 点云 计算机图形学(图像) 点(几何) 校准 人工智能 适应(眼睛) 数学 光学 物理 统计 几何学
作者
Magdalena Żuk,Marcin Majak,Izabela Szczur,Michał Popek,Ewelina Świątek‐Najwer
出处
期刊:Acta of Bioengineering and Biomechanics [Politechnika Wroclawska Oficyna Wydawnicza]
卷期号:25 (1)
标识
DOI:10.37190/abb-02296-2023-02
摘要

Purpose The objectives of the current study were to propose a calibration method for accurate augmented reality visualisation using an optical see-through head-mounted display, and to qualitatively evaluate visualisation accuracy for the application in computer assisted surgery. Methods An adaptation of stereo single-point active alignment method was proposed as the calibration procedure together with verification Three tests were performed: display of points in 3D space, on the plane, and on the surface of a skull phantom on inexperience users ( 5 participants) and experienced users ( 17 participants). Results The highest accuracy of visualization was obtained for skull phantom visualisation for an inexperienced user (3.00 mm, std 0.75 mm), while the lowest accuracy was obtained in a 3D visualisation test for an inexperienced user (22.95 mm, std 18.04 mm). The largest error was related to the depth component and amounted to 18.49 mm, std 18.10 mm for an inexperienced user. Conclusions It is possible to achieve relatively high visualisation accuracy (less than 5 mm for visualisation in space) for selected users but providing it for the group of inexperienced users seems to remain a major challenge. The accuracy of point indication can be substantially and statistically significantly increased by visualising objects on surfaces. The proposed methods and obtained results can serve as a basis for further implementation of augmented reality visualisation on an optical see-through head-mounted display in applications requiring high-quality augmented reality guidance of manual tasks.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
刚刚
lqq发布了新的文献求助10
刚刚
萤_完成签到,获得积分10
1秒前
李锟发布了新的文献求助10
1秒前
充电宝应助陈伟采纳,获得10
1秒前
烟花应助龙达绝澈采纳,获得10
2秒前
4秒前
CipherSage应助小涛采纳,获得10
4秒前
机灵难胜发布了新的文献求助10
4秒前
搜集达人应助xin采纳,获得10
4秒前
执念发布了新的文献求助10
5秒前
大佛完成签到,获得积分10
5秒前
5秒前
mmj完成签到,获得积分10
5秒前
6秒前
6秒前
没得完成签到,获得积分10
6秒前
Ava应助Camellia采纳,获得10
7秒前
7秒前
8秒前
周小凡完成签到,获得积分10
9秒前
10秒前
奥利奥发布了新的文献求助10
11秒前
11秒前
11秒前
11秒前
SHIFAN发布了新的文献求助10
11秒前
Pami发布了新的文献求助10
11秒前
mmj发布了新的文献求助10
12秒前
漂亮白云发布了新的文献求助10
12秒前
12秒前
13秒前
淡定的傲玉应助Pami采纳,获得10
13秒前
xh发布了新的文献求助10
15秒前
15秒前
wxx发布了新的文献求助10
15秒前
乐乐应助fengzi151采纳,获得10
15秒前
Eddy发布了新的文献求助10
15秒前
15秒前
松园112发布了新的文献求助30
15秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Encyclopedia of Cardiovascular Research and Medicine(2e) 820
自動車の空力技術 800
Essentials of Carbohydrate Chemistry and Biochemistry, 4th Edition 800
Organizational Behavior 510
Management and the Arts 510
Matrix Methods in Data Mining and Pattern Recognition Second Edition 510
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 化学工程 工程类 有机化学 物理 复合材料 生物化学 内科学 细胞生物学 基因 遗传学 免疫学 冶金 光电子学 癌症研究
热门帖子
关注 科研通微信公众号,转发送积分 7781325
求助须知:如何正确求助?哪些是违规求助? 9321119
关于积分的说明 20381559
捐赠科研通 7369042
什么是DOI,文献DOI怎么找? 3320035
关于科研通互助平台的介绍 2467813
邀请新用户注册赠送积分活动 2335891