PLANT SPECIES CLASSIFICATION USING EFFICIENT NET MODEL

Abstract

House plants not only beautify indoor spaces but also enhance air quality and promote mental well-being. However, identifying different species can be tricky for non-experts since many plants look alike. To address this, our study focuses on building an intelligent system that can automatically recognize various house plant species using machine learning. We trained our model on the House Plant Species Dataset, which includes about 14,790 images from 47 plant categories. To improve the system’s accuracy, we applied image preprocessing and augmentation techniques that make the model more robust. Using a Convolutional Neural Network (CNN) along with transfer learning models like ResNet50, the system learns deep visual features to accurately classify each plant. The final model can be easily integrated into mobile or web applications, making plant identification quick and accessible while supporting smart gardening and biodiversity awareness.