Development of nutrient sensor for potato plants based on foliar spectral reflectance

Developing a sensor to detect nutrients in plants, replacing wet chemical analysis, has many advantages including rapid understanding of the nutrient status and the ability to scan wide areas of the field. The sensing principle is based on detecting spectral patterns of the leaves and relating them to nutrient availability in the petiole. However, such development is complex because of the many factors (bitotic and abiotic) that may influence the status of the plants leaves and the way they reflect light. In the first stage of this research, we established a robust method of dataset creation to minimize effects of factors other than nutrients. In the second stage, we are analyzing the datasets to find correlations. At the beginning, we assumed independent relationships of the nutrients and created simple statistical models to find trends. Now, we are aiming at understanding the complex relationship between the nutrients themselves and find how they may affect each other. The results of this work will help in inventing a reliable sensor that can detect wide range of elements of nutrients.

Faculty Supervisor:

Ahmad Al-Mallahi

Student:

Partner:

Aston University

Discipline:

Engineering

Sector:

Technology; Agriculture and Food; Artificial Intelligence

University:

Dalhousie University

Program: