Hibiscus Flower Health Detection to Produce Oil Using Convolution Neural Network

Abstract

Flower market is taking a big leap in India as production and marketing is incorporating modern technologies and process improvements Flower Farming aka floriculture is facing the problem of post-harvest losses, up to 20 to 35% and forcing to address the solution concerns of farmers in this field. The focus is to classify the product, based on its health till infected condition. Farmers can deliver it to appropriate customers' leads to minimize the losses and increasing efficiency. This paper focuses on detection of healthy hibiscus flower along with its oil extraction process. The fresh flower plays especially key role to produce export quality hibiscus oil. The clear intent is to track the lifespan of hibiscus flower, where judgment of health of the flowers are addressed along with identification of infectious flowers (if detected) by applying Machine Learning Techniques like image classification and deep learning.