Image processing goals and tasks

Loading...

Del curso dictado por National Research University Higher School of Economics

Deep Learning in Computer Vision

61 calificaciones

National Research University Higher School of Economics

Deep Learning in Computer Vision

61 calificaciones

Curso 5 de 7 en SpecializationAdvanced Machine Learning

Deep learning added a huge boost to the already rapidly developing field of computer vision. With deep learning, a lot of new applications of computer vision techniques have been introduced and are now becoming parts of our everyday lives. These include face recognition and indexing, photo stylization or machine vision in self-driving cars. The goal of this course is to introduce students to computer vision, starting from basics and then turning to more modern deep learning models. We will cover both image and video recognition, including image classification and annotation, object recognition and image search, various object detection techniques, motion estimation, object tracking in video, human action recognition, and finally image stylization, editing and new image generation. In course project, students will learn how to build face recognition and manipulation system to understand the internal mechanics of this technology, probably the most renown and oftenly demonstrated in movies and TV-shows example of computer vision and AI.

De la lección

Introduction to image processing and computer vision

Welcome to the "Deep Learning for Computer Vision“ course! In the first introductory week, you'll learn about the purpose of computer vision, digital images, and operations that can be applied to them, like brightness and contrast correction, convolution and linear filtering. These simple image processing methods solve as building blocks for all the deep learning employed in the field of computer vision. Let’s get started!

Image processing goals and tasks2:51

Contrast and brightness correction5:18

Image convolution7:09

Edge detection8:30

Conoce a los instructores

Anton Konushin
Senior Lecturer
HSE Faculty of Computer Science
Alexey Artemov
Senior Lecturer
HSE Faculty of Computer Science

Now we can move to basic image processing task.

Why do we need to process images?

There are several reasons for this.

First is, image improvement for human perception to make the image better looking,

to improve the subjective image quality.

The second goal is,

to simplify the subsequent image analysis and recognition.

So, it is image improvement from machine perception.

Third goal is, image transformation for technical purposes.

For example, when we want to display video on a mobile device,

we need usually to change image resolution and

aspect ratio without damaging the content of the video.

And the last reason is, pure entertainment.

It's visual effects to get artistic impressions from the effect itself.

There are no particular reasons for these effects,

but they can be fun.

And a lot of cool applications for

mobile devices exist that apply visual effects to images.

Where are the sources of errors and image defect?

Let's look at the image transfer pipeline from

real world objects to its presentations on the computer display.

First, optical image is formed in the camera.

Then, it is passed through the wire pattern to the sensory,

where it is discretized.

Then, it is transferred via computer network and displayed on the display.

In all stages in which data is transformed and during each transformation,

there can be errors and edit artifacts which we should process.

I will give several examples of

image defects that can be compensated by image processing.

First, images can be of low contrast with

detail difficult to see in bright or dark areas.

Second, color tone can look wrong.

This picture should be more brownish.

In this slide, it is more in blue tones, which is unrealistic.

Images can be noisy or blurry because of sharp details.

Lighting in captured documents can be non-uniform,

where part of the text is very bright and part of the text is very dark.

So it's very difficult to recognize such type of document.

Image processing can also be used to extract meaningful information from images.

One example, I will discuss later, is edge detection.

Edges contain a lot of interesting information

regarding images and can be used as image features for recognition.

A lot of algorithm rely on edge detection.

Explora nuestro catálogo

Inscríbete de manera gratuita y obtén recomendaciones personalizadas, actualizaciones y ofertas.

Coursera brinda acceso universal a la mejor educación del mundo, al asociarse con las mejores universidades y organizaciones, para ofrecer cursos en línea.

© 2018 Coursera Inc. Todos los derechos reservados.

Descargar en la App Store

Consíguelo en Google Play