Machine Learning for Accurate and Efficient Pomegranate Fruit Disease Detection: A Novel Approach to Improve Crop Yield and Quality
Prashant Wakhare, S. Neduncheliyan, Pradeep B. Mane
Abstract
This paper presents a web-based tool for identifying diseases in fruit crops to address the negative impact of climate change on agriculture. With the increase in population, there is a growing demand for more food production, which puts pressure on farmers to increase crop yields. However, climate change is causing uneven weather conditions, leading to decreased crop yields. This paper proposes a solution that allows farmers to upload an image of a fruit to a system that has a trained dataset of fruit images. The system uses several image processing steps, including feature extraction based on parameters such as color, morphology, and CCV, followed by clustering using CNN Algorithm and classification using SVM to determine if the fruit is infected or not. An intent search technique is used to find the user intention from the features extracted. The experimental evaluation of the approach shows that it is effective, with a 98.38% accuracy rate in identifying diseases in fruit crops. The proposed tool can be used to identify the diseases that affect fruit crops and provide farmers with timely and accurate information, enabling them to take preventive measures to protect their crops, improve crop yields, and contribute to the overall agricultural economy. The recommendations for the application of the developed machine learning model for pomegranate fruit disease detection include using high-quality data, regularly updating the model, integrating with other detection methods, providing an easy-to-use interface, and providing technical support to growers.