Manual fruits and vegetables detection become easy when it is done in small amount, but it is a tedious process and more labor is required when gigantic amount is considered. So, automatic detection of these comes into usage. This study took the images of fruits and vegetables as input to the very first stage of processing from where detection was done. The entire process constituted three stages: Background subtraction, extraction of color as well as texture features, and then classification. Background subtraction was performed using k mean clustering technique. Color features were identified using statistical features. To identify texture features Histogram of Oriented Gradient (HOG), Local Binary Pattern (LBP) and Gray Level Co-occurrence Matrix (GLOM) were used. For training and classification, Support Vector Machine (SVM) classifier had been used and performance of this classifier had been compared with K Nearest Neighbor (KNN) classifier. After comparing the results, it shows that accuracy of SVM was higher than that of KNN. The accuracy obtained by SVM with quadratic kernel function was 94.3%.