Vision-based range estimation from a mobile robot to a target plane
Compute vision is seen as the most accurate and fastest way of doing indoor navigation. However, there must be a marker or target for the camera to trace. The marker must be in visible range so that the camera can see. If the marker is put far enough or there are some obstacles in between, computer...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/65786 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Compute vision is seen as the most accurate and fastest way of doing indoor navigation. However, there must be a marker or target for the camera to trace. The marker must be in visible range so that the camera can see. If the marker is put far enough or there are some obstacles in between, computer vision will be useless. This paper focus on WIFI positioning to overcome this disadvantage of computer vision. This dissertation is divided into two parts. The first part is WIFI positioning in long range. WIFI positioning uses the most popular and accurate finger-print method. Euclidean distance is used to evaluate the similarity and find the nearest neighbor. The second part is about short range computer vision navigation. Computer vision technique is used to make the robot be able to detect and trace the marker. OpenCV and PID controller are two major factors in this part |
|---|