没有合适的资源?快使用搜索试试~ 我知道了~
7_27am_李明杨_Visual-Inertial SLAM 理论和算法1
需积分: 0 0 下载量 173 浏览量
2022-08-03
23:55:52
上传
评论
收藏 6.76MB PDF 举报
温馨提示
试读
77页
7_27am_李明杨_Visual-Inertial SLAM 理论和算法1
资源详情
资源评论
资源推荐
Mingyang Li
2018.07.27
1
Visual-Inertial SLAM:
Theory and Algorithm
Introduction: 3D localization
Goal: high-precision localization in GPS-denied areas, using a
camera and an Inertial Measurement Unit (IMU)
2
IMU and Camera: Why
Proprioceptive sensors:
– Measure platform’s motion
– Examples: wheel encoder, inertial measurement unit (IMU)
Exteroceptive sensors:
– Capture information about the surroundings
– Examples: camera, sonar, magnetometer, laser scanner, radar
IMU + camera (vision-aided inertial navigation)
– Sensors provide 3D-motion information
– No external infrastructure required
– Small, light, low-power, and inexpensive
– Available on existing consumer devices and miniature robots
3
Applications
4
Current stage: widely used in limited number of applications
SLAM: Main Components
Estimator Design:
–
Filter based algorithm (covariance-form or information-form)
–
Iterative optimization based algorithm.
Visual measurement error function:
–
Reprojection error function.
–
Photometric error function.
Dealing with Visual-Inertial Map:
–
Marginalize all map element: Visual-Inertial Odometry.
–
Compute and keep a map online.
–
Load an offline map, and localize always against offline map.
5
You can get most state-of-the-art algorithms: By taking one
choice out of each of the three main components.
剩余76页未读,继续阅读
巴蜀明月
- 粉丝: 29
- 资源: 303
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功
评论0