Alison's New App is now available on iOS and Android! Download Now

An Introduction to Computer Vision

This free online course lays out the basics of image formation and warping and explains how computers detect features.

Publisher: NPTEL
As we open up our data to further types of inputs other than traditional text, computer vision lets machines track what is viewed onscreen. This free online course breaks down the process of analyzing image data, including object recognition and image classification, which enables us to examine our unstructured data more closely. Sign up to explore the various theoretical and practical benefits of computing with images.
An Introduction to Computer Vision
  • Duration

    3-4 Hours
  • Students

  • Accreditation






View course modules


This free online course establishes computer vision’s significance in extracting higher-level abstractions and information from images. We begin with a historical review of computer vision to track its evolution over the last few decades. We study the initial forays into computer vision and the efforts that contributed to understanding images and led to the emergence of deep learning. We also trace the origin of various designs and advances in the field of computer vision over the last half century. The course breaks down the image-formation process that uses a set of lighting conditions, scene geometry, surface properties and camera optics. We then delve into the use of digital cameras to create essential effects such as exposure, nonlinear mappings, sampling and aliasing and ‘noise’.

We then show you how to represent an image by breaking it down into tiny pixels. This includes the use of image-processing techniques to preprocess images and convert them into forms suitable for further analysis and the various image-processing operators that map pixel values from one image to another. We then unpack the method of applying filters to an image for modifying or enhancing its features. The course covers the role of histogram equalization as it uses a collection of pixel values in the vicinity of a given pixel to determine its final output value to perform local tone adjustment of the image. We demonstrate the image-enhancement process in which the transfer function of the filter modifies part of the signal frequency spectrum. We also highlight the use of Fourier analysis to determine the frequency characteristics of various filters.

The course then shows you how to use fast Fourier transform (FFT) to perform large kernel convolutions. This includes the process of decomposing an image into its sine and cosine components using Fourier transforms. We establish the significance of sampling and aliasing in the processing of images. We also demonstrate how the sampling rate determines the digitized image’s spatial resolution and the occurrence of aliasing when the sampling rate is not high enough to capture the detail in the image. Lastly, we discuss the method of increasing and decreasing the resolution of an image using interpolation and decimation filters. This informative course suits those studying computer science or anyone interested in these cutting-edge technologies.

Start Course Now