Flower Leaf Image Classification using Machine Learning Techniques

As per the report of Statista, more than 50,000 thousand categories of flower species exist worldwide; here, the problem arises identification of each type so that we can know the real advantages or the natural goodness of the flower plants. It is challenging to identify the flowers without prior knowledge/expertise. Therefore, it is crucial to make the effect and automated systems to classify the different flowers using their leaf images. This research collected 25 different categories of flowers and plants leaf images, which are 6619 total RGB images. Six classical machine learning algorithms have been utilized for the classification like K-Nearest Neighbours (KNN), Linear Regression (LR), Decision Tree (DT), Support Vector Machine (SVM), Naïve Bayes (NB), and Multilayer Perceptron (MLP). The comparative study of the classifier’s performances has been done based on classification accuracy, precision, recall, and F1-score. This research aims to find an effective machine learning classification algorithm that can be utilized for automation. The analysis of the results shows that the MLP classifier has the highest classification accuracy, i.e., 89.61%. The confusion matrix of MLP performance has been analyzed and has identified that similar shaped and textured leaves are usually misclassified.