In existing system, some tasks are manually perform such as feature extraction. Sometimes, due to manual calculation of feature extraction, the system performance is affect because each time this gave different result, which in turn reflects the accuracy rate of the system. Therefore, in the presented work, these are key point to enhance the performance of the classification system. In this research work, image classification system is present, that automatically extract the features of the images and based on the extracted features the classifier Artificial Neural Network (ANN) recognize the class of that image. For classification of images, Artificial Neural Network with the Levenberg Marquardt Training Algorithm is used. The system is implement by using MATLAB. The system gave more than 90% accuracy rate for the testing images. Therefore, based on the experimental result, it is conclude that the presented system is capable to classify the images.