The quality of the tomato is one of the most essential factors in maintaining consistent tomato fruit marketing as its maturity is the primary indicator of tomato quality in the eyes of consumers. So, an automated approach to multi-class classification for classifying various maturity stages of the tomato is proposed with the aid of various statistical features. In the proposed work, a total number of 2031 images is used to train and test the data with K-fold cross-validation (K = 10). Here, the image dataset is divided into four classes, such as Under Mature, Mature, Over Mature and Damage, based on the various stages of tomato ripening measurement. Pre-processing, feature extraction, and classification are the three major steps in the proposed approach. 10 number of statistical features obtained from each pre-processed image are used to form a feature matrix for each and every images taken from the dataset. Six machine learning algorithms, namely Support Vector Machine, Naive Bayes, Linear Discriminant Analysis, K-nearest Neighbour, Artificial Neural Network, and Random Forest are used to classify different maturity stages. In the results, it is revealed that the proposed approach obtains the maximum accuracy of 85.43% using Support vector machine over other alternative machine learning algorithms used here.