Image Enhancement and Classification of CIFAR-10 Using Convolutional Neural Networks
S. Geetha Divya, Bhaskar Adepu, P. Kamakshi
Abstract
A new architecture for enhancement of image and image classification is developed based on the image processing techniques and Convolutional Neural Multi class Networking Algorithm. Classification and enhancement of images deals with the combination of two different concepts, image enhancement and image classification techniques by identifying useful information from the given input images. This method works on image quality assessment and improving the images for identifying the image class from the dataset. The enhanced image is classified to identify the class of input image from the CIFAR-10 dataset. CIFAR-10 dataset is used to train Convolutional neural network model with the enhanced image for classification. This dataset consists of ten classes like airplane, automobiles, cat, dog, frog, horse, ship, bird, truck in colored images. This dataset is used for training and testing of our model. CNN algorithm acts as a mid-way between the image processing and image classification in the network and it is a systematic hierarchy of analyzing the images and image operations on 32*32 images.