Data Mining Techniques for Knowledge Discovery in Digital Images

Abstract

Advances in image acquisition and storage technology have led to tremendous growth in very large and detailed image databases. A vast amount of image data such as satellite images, medical images, and digital photographs are generated every day. These images if analyzed, can reveal useful information to human users. Unfortunately, it is difficult or even impossible for a human to discover the underlying knowledge and patterns in the image when handling a large collection of images. Image mining is rapidly gaining attention among researchers in the field of data mining, information retrieval, and multimedia databases because of its potential in discovering useful image patterns that may push the various research fields to new frontiers. This thesis proposes a system for classification of the images by one popular type of machine learning models which is deep neural networks, where stacked layers of “neurons” are used to learn approximate representations of data the Convolutional Neural Network (CNN). This will be done in the following suggested basic steps. The first step is to use "CIFAR-10” dataset and prepare the data for a convolution neural network. The second step is insert images to the convolution layer, Relu function, pooling layer, and flatting layer. Then using the result of the second step to classify image using fully connected and softmax function. The effectivity of using convolution neural network to classification data for train accuracy (99%) and test accuracy (95%) with bach size (128) epoch (140). The proposed approach has been applied and tested on datasets "CIFAR-10”. Using 60000 images that split into three groups. First one contains 35000 images to train the model and the second contains 15000 images to validate model while the last one is 10000 images to test the model. The implementation of the propos system has done by using python programming language