International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 07 | Jul 2023
p-ISSN: 2395-0072
www.irjet.net
Classification of Images Using CNN Model and its Variants Narayan Dhamala1, Krishna Prasad Acharya2 1Teaching Assistant, Department of Computer Science & Application, Mechi Multiple Campus, Jhapa, Nepal
2 Assistant Professor, Department of Computer Science & Application, Mechi Multiple Campus, Jhapa, Nepal
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Image classification is a method of assigning a
Multi-Layer Perceptron (MLP) also called fully connected neural networks. In such networks, each layer in one layer is connected to every neuron in next layer. CNNs can form complex patterns from simpler ones which makes them more effective in recognition of images with lots of features. In this CNN architecture, the three different layers are present which are: convolution layer, pooling layer and fully connected layer.
label to an image and it is suitable to use deep learning for this task due to spatial nature of image which can leverage the massively parallel structure to learn various features. In this research, a Convolution Neural Networks (CNN) model is presented with three configurations. The first configuration is simple and other two configurations are improvement of first configuration by using techniques to prevent over fitting. The training and testing is performed on a CIFAR-10 dataset which consists of 60000 sets of images of 10 different objects. During comparison of variants of model using different performance matrices it is observed that the dropout regularization technique can significantly make the model more accurate. It also shows that lower batch size can give better result than higher batch size. Key Words: Convolution Neural Network, Epos, Pooling, relu, softmax.
1.INTRODUCTION
Figure 1: Basic architecture of a CNN
Image classification is considered as an intelligent task of classifying images into different classes of given objects based on features. The classification problem can be binary or multiple classifications. Examples of binary classification are classifying between cat or dog images, absence or presence of cancer cells in the medical images etc. Similarly, multiple classifications include classifying cat or dog images, different animals, digit recognition etc. Image classification is used in the field of computer vision for analyzing various image data to get useful insight. It can differentiate between the given images based on tiny details which could be missed even by expert humans in the given domain It is often misunderstood with object recognition. Object recognition is a boarder term which is a combination of computer vision tasks including image classification to detect and recognize different objects in the image. So, the main difference is that image classification only deals with classifying images into different types or classes while object recognition involves detection of various objects in the image and recognizing them. Image classification involves training the machine learning model using large data sets of images. The model learns the pattern and features present in various classes of objects and can predict the class of object from previously unseen image.
1.1.1 Convolutional layer It is first and most significant layer. This layer is responsible for learning features from the input image. It takes image as input and applies a kernel (filter) to an image and produces the output. This operation is called convolution. It helps to reduce shape of the image while retaining its features. Different filters can be stacked together to extract many features. In CNN, an image with a shape (no. of images) x (image height) x (image width) is passed through Convolution layer which in turn produces a feature map of shape (no. of images) x (feature map height) x (feature map width) x (feature map channels).
1.1 Convolution Neural Network (CNN) CNN is a type of Deep Neural Networks which is mainly used for solving image recognition related problems. It a variant of
© 2023, IRJET
|
Impact Factor value: 8.226
Figure 2: Convolution operation
|
ISO 9001:2008 Certified Journal
|
Page 1062