Plant Disease Detection System

It is apparent that the agriculture sector plays a critical role within the Vietnamese economy. Therefore, it is essential to maintain agricultural productivity through sustainable means while guaranteeing food security on a national scale. This project aims to develop a non-invasive method that detects plant disease during the initial stages of development. By identifying early signs of the disease, less reliance on harmful antibiotics or pesticides will be used, leading to higher food production levels and quality. 



Within this project, the system comprises of two main parts: Computer Vision and Camera System. The Camera System consists of a hyperspectral camera that scans and produces image data of a plant in which the belt driven system and an optical camera control in a linear motion. Afterwards, the image data transfers to the main computer to perform preliminary wavelength processing to filter out noise. The team processes the remaining crucial data for any signs of disease known as Powdery Mildew within the Computer Vision System. The current prototype proves to be effective in identifying patterns of the Powdery Mildew disease based on the hyperspectral data and further demonstration will be presented using the testing frame with the integrated system.