Flower classification is a crucial task for understanding biodiversity, tracking climate changes, and protecting endangered plants. In this paper, we propose a deep learning approach using a convolutional neural network (CNN) architecture for accurate and efficient flower classification. Our methodology includes preprocessing the dataset, implementing the CNN architecture, and training the model using stochastic gradient descent with cross-entropy loss. Our results demonstrate that our approach achieves an accuracy of 91.73% on the test set, which is comparable to or better than other sophisticated models. Ablation studies reveal the importance of each component of our CNN architecture, while our data preprocessing step improves the model’s generalization performance and prevents overfitting. Our study provides a reliable and effective deep learning approach for flower classification that can be used in various applications, including botany, agriculture, and ecology.