Feature Extraction and Classification Performance of Fine-Tuned ResNet-50 in Plant Species Recognition

This work proposes a fine-tuned ResNet-50 model-based floral classification system using its strong feature extraction capacity to improve classification accuracy. ResNet-50, a deep convolutional neural network pre-trained on ImageNet, shines in capturing complex patterns because of its residual learning methodology, which helps to build deeper networks. Fine-tuned ResNet-50 shows superior performance in managing demanding picture datasets than standard models since it lowers vanishing gradient difficulties and allows effective learning. The dataset consists of two, 746 photographs in five classes: Daisy (501 images), Dandelion (646 images), Rose (497 images), Sunflower (495 images), Tulip (607 images). Good evaluation and training are made possible in part by this fair sharing. Following lengthy training and validation, the model shown remarkable accuracy of 91% thus proving its practical significance in plant identification and biodiversity monitoring. The results show how effectively deep learning techniques—especially enhanced ResNet-50—in advance of floral classification problems work. Better classification accuracy and efficiency in plant species identification help to enhance ecosystem management by means of which sustainable development goals (SDGs) are achieved and agricultural output is raised.