Introduction to Content-Based Image Classification
Rik Kamal Kumar Das · 2020
Image data is a significant source of information to leverage data-driven decision making. The rich content embraced in images can divulge valuable information that is useful for assorted types of applications. Today, we witness manifold usage of image content in various disciplines, including remote sensing, medical imaging, media, etc. The current generation is getting increasingly familiar with image-based conversations. The fact can be well-supported by visualizing a plethora of photo uploads in social networking websites. Maintenance, archival, storage and utilization of these data in real time are a challenging task. Content-based image classification (CBIC) has addressed this bottleneck efficiently by categorizing images into subsequent classes on the basis of image content. The process considers significant image content-based features as identifiers to recognize the image category. This is in contrast to traditional text-based image classification, which relies on manual annotation of image data. The popularity of machine learning and deep learning techniques has further enhanced the acceptability of content-based image classification in the contemporary period. This chapter provides an overview of content-based image classification in terms of its background and motivation. It also briefs you about various classification techniques used in this book to illustrate content-based image classification along with the evaluation metrics. The chapter mentions the open datasets used for experimentation along with the specifications of images in those datasets. Finally, it furnishes a concise description of the organization of the book.