Nowadays it is vital to automatically identify and recognise medicinal plant species in settings like forests, mountains, and dense areas in order to be aware of their existence. The shape, geometry, and texture of various plant parts, such as leaves, stems, flowers, etc., are now used to identify plant species. Systems for identifying plant species based on their flowers are frequently employed. Even though contemporary search engines offer tools for visually searching for a query image that includes flowers, these methods lack robustness due to the intra-class variance among the millions of flower species found worldwide. Therefore, a Deep Learning technique using Convolutional Neural Networks (CNN) is applied in this proposed study work to accurately identify flower species. The cellular phone’s built-in camera module is used to capture images of the different plant species. A Transfer Learning technique is used to extract complicated characteristics from pre-trained networks for floral image feature extraction. Generally, to increase accuracy, a machine learning classifier like Logistic Regression or Random Forest is added on top of it. This strategy aids in reducing the amount of hardware required to complete the computationally demanding task of training a CNN. It has been found that employing the VGG-19 pre-trained model architecture and CNN paired with Transfer Learning methodology as a feature extractor outperforms all manually created feature extraction techniques like Local Binary Pattern (LBP).