Design of Image Classifier for Web Application

An image classifier uses machine learning algorithm to recognize the images. A classifier allocates class labels to specific data points and classifies the images using supervised learning. It recognizes the target classes using labelled sample images by training a Convolutional neural network model (CNN). This paper aims to design an image classifier for web application. Image classification will eliminate the need of manual search in the web application. CNN is an efficient algorithm to classify the images according to the category label. An image classification model has been designed using CNN to classify the images of fruits and vegetables. Accuracy of the model is based on the classification technique. For successful classification data pre-processing, data augmentation and feature extraction are involved. Positive results were obtained by the CNN model after testing it with the input images. The model can classify the 90483 images of 131 fruits and vegetables. This paper involves the description and methodology for designing the image classifier.