Citation

Muliady, Muliady and Lim, Tien Sze and Koo, Voon Chet (2023) Machine Learning Based Algorithm for Rice Leaf Nitrogen Status Estimation. In: 1st FET PG Engineering Colloquium Proceedings 2023, 16 June - 15 July 2023, Multimedia University, Malaysia. (Submitted)

Text 22_PPM Poster_CRSST_Muliady_June2023.pdf Restricted to Registered users only Download (420kB)

Abstract

Rice plant site-specific nutrient management need farmer to measure Nitrogen status of rice plants. This research objective is to develop a machine-learning based algorithm to estimate the rice leaf Nitrogen status based on a single sensor multispectral images. The experiment was set by planting 49 rice plants in pots. A Red Edge and an Orange Cyan Near Infrared (OCN) Mapir Survey3 camera were used, with an additional light sensor. A total of 768 pairs of data were created which have features of NDVI value, RE value, light intensity, and time with SPAD value as output. The best performance of the Support Vector Regression (SVR) algorithm is achieved with Radial Basis Functions kernel, gamma value 8, and epsilon value 0.1. The estimation using a testing dataset gives an average error of 2.63%, which means the model can relatively estimate the Nitrogen status accurately. But in situ testing, the average error is significantly increased to 23,4%. This may caused by the different environments from the training dataset, further experiment is still needed to create an in situ dataset for better performance.

Item Type:	Conference or Workshop Item (Poster)
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions:	Faculty of Engineering and Technology (FET)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	15 Aug 2023 01:42
Last Modified:	15 Aug 2023 01:42
URII:	http://shdl.mmu.edu.my/id/eprint/11621

Downloads

Edit (login required)