skip to main content

ACCURACY EVALUATION OF OIL PALM TREE COUNT USING OBIA AT TENERA SUMBER REJEKI FARMERS GROUP PLANTATION, TERANGKEH, KOTABARU, SOUTH KALIMANTAN

Department of Earth Technology, Vocational School, Gadjah Mada University, Indonesia 55288, Indonesia

Received: 23 Apr 2026; Published: 20 Oct 2026.

Citation Format:
Abstract

Indonesia holds the largest oil palm plantation area in the world, making accurate tree count data essential for plantation management. Conventional field survey methods are limited in time and cost efficiency, making them impractical for large-scale areas. This final project aims to automatically count oil palm trees using the Object Based Image Analysis (OBIA) method, evaluate detection accuracy, and produce a tree distribution map at the Tenera Sumber Rejeki Farmer Group Plantation, Kotabaru Regency, South Kalimantan. Secondary data were obtained from PT Eshan Agro Sentosa, comprising aerial photo processing results and field survey records. Data processing was conducted using eCognition Developer and QGIS through segmentation and classification stages. Segmentation using the Multiresolution Segmentation algorithm yielded optimal parameters: Scale Parameter (200), Shape (0.3), and Compactness (0.7). Classification was performed using the Nearest Neighbor method with Cluster Random Sampling across five object classes. The OBIA method detected 1,479 oil palm trees within a 10.51-hectare study area, compared to 1,521 trees in the field survey. Accuracy assessment yielded a Detection Percentage (DP) of 93.17% and Branch Factor (BF) of 5.88%, both classified as excellent. Results were visualized in a distribution map integrating OBIA outputs with field survey data.

Article Metrics:

Article Info
Section: Articles
Language : EN

Last update:

No citation recorded.

Last update:

No citation recorded.