Accurate plant identification remains a significant challenge in the field of computer vision due to species variability, similar physical features of plants and leaves, environmental conditions, lighting variations, and limited annotated datasets. However, traditional methods are very dependent on expert botanists and manual classification, often taking a lot of time and being prone to errors. In this study, an automated plant detection system is developed that is trained using a large (over 10,000 images) high-resolution dataset of images to detect medicinal plants in the form of *Azadirachta indica* (Neem) and *Bacopa monnieri* (Brahmi). As a result, the model achieves a precision of 0.97 while using extensive data augmentation techniques and deep learning-based object detection to ensure species are accurately classified despite problems such as species similarity, environmental variability, and lighting inconsistencies. The system works well in the real world under conditions of occlusion, seasonal changes, and dataset limitations. The experimental results show that the model exhibits robust classifier performance in distinguishing closely related species with a small margin for error.