Implementation of Virtual Instrumentation System for Estimation of Eaten Leaf Area using Digital Image Processing

There are many reasons why a plant’s yield can be lost. One of these is due to the presence of diseases and pests. The complexity of identifying these diseases and their prevalence can be challenging for a farmworker. If proper care is not taken in this area, it can cause various problems for the plants and industries. This can affect the quality of the products and the productivity of the plants. This paper presents a Virtual Instrumentation system for detecting and estimating the eaten leaf area of plants. It uses Otsu thresholding, cluster-based thresholding, entropy, metric, and moments algorithms on a self-generated dataset of 500 images of healthy and unhealthy leaves. The model achieves its best accuracy of 98.20% through Otsu’s thresholding algorithm. The proposed algorithm is a generalized approach that can be used for any plant. It is a computationally cheap and convincing method. This research would support food processing industries, particularly the sorting of leafy green vegetables, where the main objective is to pack a high-quality product free of all defects.