10 - Active Contours - Introduction with ipol.im and Photoshop Demos - Duration 05:58

Loading...

Do curso por Duke University

Image and Video Processing: From Mars to Hollywood with a Stop at the Hospital

371 classificações

Duke University

Image and Video Processing: From Mars to Hollywood with a Stop at the Hospital

371 classificações

In this course, you will learn the science behind how digital images and video are made, altered, stored, and used. We will look at the vast world of digital imaging, from how computers and digital cameras form images to how digital special effects are used in Hollywood movies to how the Mars Rover was able to send photographs across millions of miles of space. The course starts by looking at how the human visual system works and then teaches you about the engineering, mathematics, and computer science that makes digital images work. You will learn the basic algorithms used for adjusting images, explore JPEG and MPEG standards for encoding and compressing video images, and go on to learn about image segmentation, noise removal and filtering. Finally, we will end with image processing techniques used in medicine. This course consists of 7 basic modules and 2 bonus (non-graded) modules. There are optional MATLAB exercises; learners will have access to MATLAB Online for the course duration. Each module is independent, so you can follow your interests.

Na lição

Image segmentation

Not all parts of the image are the same, and students will learn the basic techniques to partition an image, from simple threshold to more advanced graph cuts and active contours. This is the first unit where student will learn about image analysis and image interpretation, and will learn why this is important, e.g., in medical imaging and object recognition.

1 - Introduction to Segmentation - Duration 04:174:17

2 - On Edges and Regions - Duration 05:175:17

3 - Hough Transform with Matlab Demo - Duration 20:5920:59

4 - Line Segment Detector with Demo - Duration 03:203:20

5 - Otsu's Segmentation with Demo - Duration 14:2514:25

6 - Congratulations - Duration 00:170:17

7 - Interactive Image Segmentation - Duration 21:1321:13

8 - Graph Cuts and Ms Office - Duration 09:349:34

9 - Mumford-Shah - Duration 05:505:50

10 - Active Contours - Introduction with ipol.im and Photoshop Demos - Duration 05:585:58

11 - Behind the Scenes of Adobe's Roto Brush - Duration 31:29 - Optional breaks at 20:30 and 27:2631:29

12 - End of the Week - Duration 00:210:21

Conheça os instrutores

Guillermo Sapiro
Professor
Electrical and Computer Engineering

Hello and welcome back. Active contours, also called snakes, are

very important in image and video processing.

They appear in a number of commercial products and they are very useful for

numerous image and video segmentation. Let me illustrate the basic concept

behind active contours. The idea is that you start from one or

multiple curves. For example, I'm going to draw a curve

that surrounds the flower, which is my object of interest.

And the basic idea behind active contours is to design velocities that will move

this curve towards the flower. Hopefully, as the curve moves, it's going

to end up in the border of the flower. So, we need basically to design

velocities that will move any initial contour towards the contour of interest,

which is the boundary of the object of interest.

Now, I can write down, the formulation for this and the equations for this, but

I think it's better if we first learn the material that is coming next week, as a

back ground, and that will help us to write this in a much, much elegan, form.

So, this week, instead of writing down the exact weight

of active counters, I'm going to just illustrate you some

examples to motivate you and get you very excited to learn the background, material

that we're going to learn next week. So, let's go and watch some movies with

examples of active contour, inside the image processing online

package and inside Adobe's Photoshop. So, let's just observe that right now.

In order to illustrate one of the examples of active contours, we're going

to be using, again, the image processing online website where, as we have seen

before, there is many very interesting image processing algorithms implemented.

And particularly, now, we are in the Chan-Vese webpage,

which is an implementation of one of the approaches for active contours developed

by Tony Chan and Luminita Vese. Before we show you one of the examples

running actually in real time, let's just basically run the demo provided by the

author of this website. So here, we see contours evolving.

This is a very clear example of a contour evolving towards the segment.

Here, we have another one, just look at this one.

And here, we see once again, how contours are evolving and yet once again.

So, that's a basic idea of active contour. This was a very fast video, just

to illustrate one of the examples. But let's just run one, so we're going to

have more time, to look at it. Wwe're going to run the demo online as we have

done in the past for other algorithms. I'm going to pick this, example, it's

just one example, I'm not going to change the parameters, we don't know exactly

since we haven't discussed the technique, what this parameter means, so I'm going

to just, leave the default and I'm going to run it.

Remember, now is rhyming as we see here, but it's rhyming in the server wrap that

is provided by the imaging processing online.

The algorithm is actually much, much faster, than we can see, it actually can

be run on real time, but it's run on the server.

And here is the result. We see once again, the initial contours,

these intial contours at the very beginning are the initial contours and

then, we see the boundaries. The curves have evolved toward those

boundaries. The curves have evolved to segment out

the object of interest. There are some spurious objects and we

are going to understand why they happen, but mostly, it has done a great job

basically, in finding the boundary of the flower, which is probably the object of

interest for this image. And we see once again, these contours

evolving. And this is, as I say, something that

we're going to learn next week. Exactly the formulas, exactly the math

behind it. But this illustrates the idea.

Thank you for this part of the video. It's now time to illustrate using Adobe

Photoshop one implementation of active contours.

In particular, we're going to be using this tool in active contours that is a

selection tool and look what's going to happen.

Please observe my mouse. Look how I'm moving the mouse and the

contours are being automatically attracted to the boundary of the flower.

You know, how I just clicked and it's basically catching the flower.

I almost didn't move my mouse, I just click inside the object of interest, and

we capture the flower. Let me do that again here. I'm going to

just click and it's capturing this part of the flower.

So, you see how we can help, and we have already learned about user interaction by

doing a few clicks inside the flower. Basically, a technique related to active

contours, move my clicks towards the boundaries of the object of interest.

This is a very, very interesting technique,

this time, implemented inside Adobe Photoshop.

And this provides in our demo of active contour type of approaches.

Thank you. These examples that we just saw should

motivate you for the material that is coming next week.

Before we go into that, we have a couple of additional and very, very exciting

things to learn about video segmentation. So, I see you in the next video.

Thank you.

Explore nosso catálogo

Registre-se gratuitamente e obtenha recomendações, atualizações e ofertas personalizadas.

O Coursera proporciona acesso universal à melhor educação do mundo fazendo parcerias com as melhores universidades e organizações para oferecer cursos on-line.

© 2018 Coursera Inc. Todos os direitos reservados.

Baixar na App Store

Baixar no Google Play