A low cost aerial remote sensing platform for counting oil palm trees

Md Noor, Mohd Faris Murshidi (2019) A low cost aerial remote sensing platform for counting oil palm trees. [Project Paper] (Submitted)

Text lp FK 2019 81.pdf Download (33MB)

Abstract

Unmanned Aerial Vehicle (UAV) System has gained popularity in the field of photogrammetry, remote sensing and geospatial engineering for civilian application. Its impact on society and the speed of project delivery has undeniably present users with varieties of choice and producers with very attractive competitiveness. In large scale or area for vegetation and agricultural application, satellite remote sensing has been widely used for collecting the data. However for small to medium scale enterprise or medium farm, UAV very idea to use for space and airborne platforms are very economical due to budget constraints. In this study, an RGB image was captured by UAV and used to estimate the palm tree in the interested area. To estimate trees inside the interested area, Training Rule Method was used. The value of the interested area and the factor of area were multiplied by number of trees in the training area. This method was also compared to the Trimbler eCognition Developer software, in order to compare the accuracy of trees estimated. On eCognition Developer the values of threshold used were (0.5) and (0.4). The Training Rule method gives the result of first image 94% and second image 96% of accuracy. Meanwhile, eCognition method gives the result of threshold (0.5), on first image 92% and second image 85% of accuracy. For threshold (0.4), on first image 95% and second image 99% of accuracy.

Item Type:	Project Paper
Subjects:	T Technology > TA Engineering (General). Civil engineering (General)
Faculty:	Faculty of Engineering
Depositing User:	Ms. AZLINA ZAINAL ABIDIN
Date Deposited:	26 Nov 2025 07:24
Last Modified:	26 Nov 2025 07:24
URI:	http://psaspb.upm.edu.my/id/eprint/2627

Actions (login required)

View Item