Assessment of automated road features xtraction algorithm from UAV images / Amirul Ahmad
In these days, thorough documentation of the road network is vital. It is especially true for many applications such as managing transportation and automation of navigation. Therefore, the extraction of road network such as from Unmanned Aerial Vehicle (UAV) imagery is needed so that it can be made...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/43728/1/43728.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In these days, thorough documentation of the road network is vital. It is especially true for many applications such as managing transportation and automation of navigation. Therefore, the extraction of road network such as from Unmanned Aerial Vehicle (UAV) imagery is needed so that it can be made use for these applications. The road network extraction can be done manually, however, it is costly and time consuming to update and utilize the spatial information compared to automatic extraction. The aim of this study is to analyze the capabilities of automatic road extraction from UAV images using Trainable Weka Segmentation (TWS), Level Set (LS) and Seeded Region Growing (SRG) method. To achieve this, the objectives of this study are to: 1) extract road automatically using TWS, LS and SRG method and 2) examine the capabilities of automatic road extraction from UAV images. The study area was carried out at UiTM Arau, Perlis, Malaysia. To ensure the completion of all objectives, several Ground Control Points (GCPs) had been established at UiTM Arau. Lastly, Agisoft PhotoScan had been used to build the orthophoto which then the road network in the orthophoto had been segmented and extracted using these ImageJ Fiji. The automatic extracted road network had then been compared to manually extracted road network. It was found that SRG method is slightly better in extracting road features compared to LS method. This study can help reducing the cost and time consumed in extracting features, especially road network, by using automatic extraction instead of manual extraction. |
|---|