Identifikasi Jenis Pohon Dengan Aerial Laser Scanner LiDAR (Studi Kasus: Kawasan Wisata Waduk Selorejo)

Oktavia, Hana´ Nafisa (2024) Identifikasi Jenis Pohon Dengan Aerial Laser Scanner LiDAR (Studi Kasus: Kawasan Wisata Waduk Selorejo). Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5016201111-Undergraduate_Thesis.pdf] Text
5016201111-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2026.

Download (8MB) | Request a copy

Abstract

Menurut Kementerian Lingkungan Hidup dan Kehutanan (Kemenlhk, 2022), Indonesia memiliki kawasan hutan mencapai 125,76 juta hektar pada tahun 2022. Waduk selorejo merupakan wilayah yang termasuk dalam kawasan hutan atau vegetasi alami yang tumbuh di sekitar waduk, memberikan manfaat ekologis dan estetika yang penting bagi lingkungan sekitarnya. Perlunya dilakukan pemetaan jenis pohon sebagai upaya untuk pemantauan sumber daya hutan, penilaian keanekaragaman hayati, pengambilan biomassa, dan penyerap karbon hutan. Mengidentifikasi pohon masih dilakukan secara tradisional hingga saat ini. Identifikasi jenis pohon mengandalkan pembuatan plot sampel untuk survei manual, dengan kelemahan yaitu rentan waktu yang lama, biaya tinggi, dan intensitas kerja lapangan yang tinggi. Penelitian ini dilakukan ekstraksi data LiDAR untuk memisahkan antara objek pohon dan non-pohon (bangunan, jalan, dll) dan metode point cloud segmentation dalam proses mendeteksi dan menggambarkan individu pohon. Algoritma segmentasi yang diusulkan mengisolasi pohon secara individual dan berurutan dari titik awan, dari pohon tertinggi hingga terpendek. Kriteria yang dihasilkan berupa berupa Tree ID, Tree Position (X, Y) pada koordinat UTM, Tree Height, Crown Diameter, Crown Area, dan Crown Volume (Li et al., 2012). Selain itu, dilakukan akurasi segmentasi individu pohon dan evaluasi ground truth dengan data tinggi pohon dan jenis pohon. Metode pengolahan tersebut untuk mengidentifikasi individu pohon di wilayah sekitar Waduk Selorejo tahun 2023 didapatkan sejumlah 432 pohon tunggal. Terdapat 20 pohon yang dapat terdeteksi dengan sempurna. Diantaranya pohon tersebut berjenis pohon trembesi (samanea saman) dengan 9 individu pohon, pinus (pinus canariensis) dengan 5 individu pohon, dan palem raja (roystonea regia) dengan 2 individu pohon. Didapatkan akurasi segmentasi individu pohon memiliki nilai akurasi sebesar 0,037.Didapatkan RMSE total memiliki nilai yang cukup besar dengan nilai 0,913 meter. Perbedaan ini cukup jauh dikarenakan kemungkinan terdapat beberapa faktor yaitu rentan waktu data LiDAR dan data ground truth dan persepsi data ground truth dengan LiDAR dalam pengambilan tinggi pohon. Dapat ditemukan vegetasi medium sejumlah 1.453.693 point cloud dan vegetasi tinggi sejumlah 1.654.228 point cloud.
==================================================================================================================================
According to the Ministry of Environment and Forestry (Kemenlhk, 2022), Indonesia has a forest area of 125.76 million hectares in 2022. Selorejo reservoir is an area included in the forest area or natural vegetation that grows around the reservoir, providing important ecological and aesthetic benefits to the surrounding environment. Tree species mapping is necessary for forest resource monitoring, biodiversity assessment, biomass collection, and forest carbon sequestration. Identifying trees is still done traditionally today. Tree species identification relies on the creation of sample plots for manual surveys, with the disadvantages of being time consuming, high cost, and high field work intensity. This research conducted LiDAR data extraction to separate between tree and non-tree objects (buildings, roads, etc.) and point cloud segmentation method in the process of detecting and delineating individual trees. The proposed segmentation algorithm isolates trees individually and sequentially from the point cloud, from the tallest tree to the shortest. The resulting criteria are Tree ID, Tree Position (X, Y) in UTM coordinates, Tree Height, Crown Diameter, Crown Area, and Crown Volume (Li et al., 2012). In addition, individual tree segmentation accuracy and ground truth evaluation with tree height and tree species data were performed. The processing method to identify individual trees in the area around Selorejo Reservoir in 2023 obtained a total of 432 single trees. There are 20 trees that can be detected perfectly. Among them are trembesi (samanea saman) with 9 individual trees, pine (pinus canariensis) with 5 individual trees, and king palm (roystonea regia) with 2 individual trees. The segmentation accuracy of individual trees has an accuracy value of 0.037. The total RMSE obtained has a fairly large value with a value of 0.913 meters. This difference is quite far because there may be several factors, namely the time vulnerability of LiDAR data and ground truth data and the perception of ground truth data with LiDAR in taking tree height. Medium vegetation can be found in the amount of 1,453,693 point clouds and high vegetation in the amount of 1,654,228 point clouds.

Item Type: Thesis (Other)
Uncontrolled Keywords: Segmentasi Individu Pohon, UAV LiDAR, Pohon, RMSE, Individual Tree Segmentation, UAV LiDAR, Trees, RMSE
Subjects: G Geography. Anthropology. Recreation > G Geography (General) > G70.5.I4 Remote sensing
T Technology > TR Photography > TR810 Aerial photography
Divisions: Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geomatics Engineering > 29202-(S1) Undergraduate Thesis
Depositing User: Hana' Nafisa Oktavia
Date Deposited: 08 Aug 2024 06:02
Last Modified: 08 Aug 2024 06:02
URI: http://repository.its.ac.id/id/eprint/114792

Actions (login required)

View Item View Item