A Low-Cost Embedded IoT System for Real-Time Rice Leaf Nitrogen Assessment in Tropical Smallholder Agriculture
ID：121
View protection：Participant Only
Updated time：2026-10-04 23:46:15 Views：39
In-person
Abstract
Inefficient nitrogen (N) management in smallholder rice farming leads to low fertilizer use efficiency (30–50%) and environmental losses. This paper presents the design, implementation, and field validation of a low-cost (₱20,000/USD $360) embedded IoT system for real-time rice leaf N assessment under tropical field conditions in Catanduanes, Philippines. The embedded device integrates an Arduino-based microcontroller, OV7670 RGB camera with controlled 6500K LED lighting, NEO-6M GPS module (±3m accuracy), and LCD display. On-board image processing using RGB-to-HSV color space conversion estimates N status in <30 seconds per sample. Validation with 20 rice farmers (80% aged ≥51 years; mean experience 28.7±14.2 years; 70% no prior technology use) demonstrated exceptionally high demonstration effectiveness (composite mean = 4.79/5.00, SD=0.413). Overall usability was rated 4.50/5.00 (SD=0.581) on a 5-point Likert scale, with Cronbach's α = 0.87 indicating excellent internal consistency. Perceived ease of learning (4.58±0.50) and ease of use (4.55±0.54) received the highest ratings. A significant positive correlation was found between demonstration effectiveness and adoption intention (r=0.68, p<0.01). The device addresses three major leaf color chart constraints: ambient light dependency, operator subjectivity, and lack of geospatial tagging, offering a scalable embedded solution for precision N management in resource-limited tropical agriculture.
Keywords
embedded systems,Internet of Things,leaf color chart,nitrogen management,low-cost sensor,rice smallholders
Post comments