Recognition of Pistachio Species with Transfer Learning Models
Fagun Patel, Shubbh Mewada, Sheshang Degadwala, Dhairya Vyas
Abstract
This comprehensive research delves into the intricate task of recognizing diverse pistachio species, spanning applications in agriculture, ecology, and the food industry. By harnessing the power of transfer learning models, including established architectures like AlexNet, VggNet, and ResNet, alongside the novel Convolutional Neural Network (CNN) design enhances the accuracy and efficiency of pistachio species recognition. By fine-tuning pre-trained models and training the CNN from scratch, their ability is leveraged to capture both general and species-specific features. This rigorous evaluation employs an extensive dataset, embracing a wide array of pistachio images, to meticulously gauge model performance using metrics such as accuracy, precision, recall, and F1-score. This research not only sheds light on the comparative prowess of transfer learning models and this unique CNN architecture in pistachio recognition but also offers insights with implications for broader fields like agriculture and botany. Ultimately, this study underscores the potential of innovative deep learning approaches to advance species recognition accuracy while showcasing the versatility of the proposed CNN model in diverse botanical applications.