Putri, Nurul Supremasinta (2016) Ekstraksi Jalan Raya Berdasarkan Pendeteksian Zebra Cross Dari Foto Udara Beresolusi Sangat Tinggi Dan Data DSM. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
5112100197-Undergraduate_Thesis.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Terdapat beberapa strategi pada proses ekstraksi
jalan seperti segmentasi citra, deteksi garis, evolusi kurva,
multi-resolution, kecerdasan buatan, deteksi tepi, roadtracking,
operasi morfologi, pengenalan objek. Semua strategi
dari ekstraksi jalan sangat tergantung pada karakteristik data.
Foto udara dan data ketinggian menjadi data pada
tugas akhir ini. Mengintegrasikan foto udara dan data
ketinggian akan menyelesaikan kelemahan masing-masing
data dalam pengekstraksian jalan.
Pada Tugas Akhir ini digunakan pendeteksian zebra
cross sebagai langkah awal untuk mengenali jalan,
dilanjutkan dengan threshold, region growing, dan yang
terakhir, road line filtering untuk mengekstraksi jalan.
Dari hasil uji coba didapatkan hasil quality
pengekstraksian jalan yang dilakukan terbaik sebesar 42,2%.,
persentase jalan yang terdeteksi sebesar 86,5%, dan
persentase objek bukan jalan yang terdeteksi sebesar 71,1%
yang menggunakan metode threshold dan road line filtering.
==================================================================================================================
There are some strategies in term of road extraction
process such as: Image segmentation, edge detection, curve
evolution, multi-resolution, artificial intelligent, line detection,
road tracking, morphology operation, object recognition. All
strategies of road extraction are very depend on the
characteristic of the data.
RGB aerial image and DSM data be the input data in
this undergraduate thesis. Integrating the aerial image and
elevation based data will overcome the shortcomings and
weaknesses of each type of data in road extraction.
In this undergraduate thesis, to recognize the road used
zebra cross detection as initial step, and then used threshold,
region growing, and the final step is road line filtering to
extract the road.
From experimental results, the quality of road
extraction is 42,2%, percentage of detected road is 86,5%, and
percentage of detected another object (non road) is 71,1%
which used threshold and region growing method .
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSIf 526.982 Put e |
| Uncontrolled Keywords: | Foto udara; resolusi sangat tinggi; data DSM; ekstraksi jalan; pendeteksian |
| Subjects: | T Technology > TR Photography > TR267.733.M85 Multispectral imaging |
| Divisions: | Faculty of Information Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
| Depositing User: | ansi aflacha |
| Date Deposited: | 23 Apr 2020 04:25 |
| Last Modified: | 23 Apr 2020 04:25 |
| URI: | http://repository.its.ac.id/id/eprint/75891 |
Actions (login required)
 |
View Item |