Kalman Filter – A painless approach

Needless to say but Kalman Filtering is one of the most powerful estimation processes in almost any Engineering field. From robotic vacuums to Satellite Guidance, it is everywhere. Here I will explain the how’s and why’s of the Kalman Filter (KF) in our lives.


Any decent technological project will use this robust method for the final estimation of the position of any intelligent system. The format of the given information can be, fortunately, represented as a Gaussian state.

Thanks to this property, it is possible to use Gaussian filters (KF is one of them), in order to improve the final estimation.

Gaussian modelling estimations are going to be carried in this explanation, so it is preferable to have this mathematical background (a simple understanding is enough) in order to follow the presented technique. Continue reading


Visual Navigation for Flying Robots

Free lectures! Free courses on-line! Free! Free!

Got your attention? I’m like spam.

So far, so good. It has been a year since I decided to get fully involved in this field of Computer Vision, and hasn’t been easy, but I gotta say it is full of surprises. My University has been strongly doing some research about it, and a proof of that is that the 3D Computer Vision course given in Coursera is actually done by a Professor in my university (tho in German). So, I think we’re moving quite strong.


My curiosity took me to the Computer Vision chair of the Informatics Department at the TUM and I think this semester they completely got me, as I’m planning to get the 3 lectures given there: Visual Navigation for Flying Robots, Multiple View Geometry and Machine Learning for Robotics and Computer Vision. Robotics, Machine Learning and Computer Vision! Makes my mouth water. Huh! Beautiful. Continue reading

3D Reconstruction of a face from a single image

I don’t know if scary, jaw-dropping or simply impressive, but it turns out that there is already an algorithm developed here in Germany by Professors Volker Blanz and Thomas Vetter, that reconstructs a face in 3D from one image. Yes! out of a single image!

Maybe it wouldn’t sound that amazing at first sight, but you have to think about all the possible applications. Besides identifications, you can entirely play around with the information out of peoples’ faces, prediction, tracking, augmented reality, medical procedures, etc.

Is it something new? No way! It was presented in the International Conference on Computer Graphics and Interactive techniques of 1999! That’s right, fella, 14 years ago!

But why nobody used this impressive technique? Well, my guess is, of course, because of the computation times of it. Yes yes, very good, looks great and reconstructs faces pretty good, BUT took 50 minutes to reconstruct a face.

But wait a second! It was 14 years ago, right? With slower computers. Certainly, according to their paper, they used an SGI R10000 processor, which at that time was as powerful as 250 MHz… Well, I think is time to try it again with our new multi-core processors at several GHz.

Is it still on development? Did the creators give up? Are they preparing some new surprises? Well, we don’t know; but so far I can say that this technique has tons of applications with the new devices focusing in computer vision applications.

Just check the demo about this great procedure and the explanation of its development:

Predator Algorithm

The more I go into Machine Vision, the more I get amazed by the Engineering behind the most common thing around me.

Now, this post is not abut something that we can see daily now, but I’m sure it’s gonna be soon among us. Very soon. I’m talking about the Predator Algorithm (originally named TLD), developed by czech scientist Zdenek Kalal. This algorithm is implemented in such a way that the computer learns continuously detecting the features and objects in it.

Yes, it learns and with time detects easily every object. But well, nothing better than a graphical explanation of what I’m trying to explain, by the inventor himself:

The creator of this algorithm had such a success that he already started his own company to start the implementation of this great Engineering tool: TLD Vision.

These are the kind of things that really inspire.

The Fundamental Matrix song

Annoying, exciting, boring, awesome… Computer Vision and its applications might be sometimes a roller-coaster of emotions, but it is always a pleasure to work with. Some of those familiar with Vision systems, might be aware of some nice properties and most of all, some fancy tools that help us to play around with images and computers; but one important thing that no one should forget in this field is the so-called Fundamental Matrix. If you have still some doubts about its development, then here you have a bit of its application:

Looking for the lyrics. Well, I totally recommend you to go to the website of Daniel Wedge, the creator of this totally entertaining song.

Build your own 3D Scanner!

Yes, it is possible and you can do it. Actually you don’t have to be a total expert in programming or Maths. Yes yes, you have to know some good stuff, but being a nerd or a genius is not the point.

Gabriel Taubin, an Argentinian Mathematician researching in Brown University is in charge of this project and made it available for everyone, so if you have 3D Computer Vision as your hobby, then this is the perfect start. As I said, you don’t have to be an expert but you need to go deep into it and it will be totally amazing, I promise. Just have a look to the website:


The final results would be amazing and they provide every kind of tool that you may need. It seems that the used methods vary, but that’s just extra fun. I really encourage you to try it, because it’s cheap, fun and totally useful.

The following video is NOT of the project, but it shows you how accessible is to build your own 3D scanner.

But don’t forget the Kinect, boy!

My first day in Computer Vision

A new semester in the Uni has begun and in this second week I am already feeling the first pressured tasks. No probs, this time planning and hard work will be a constant besides a good health caring and exercise. Perhaps less Internet Facebook is what I need.

This semester I will struggle a bit with two very interesting courses: 3D Computer Vision and Image Understanding in English and German, respectively. Before I had “Machine Learning Method for Computer Vision Applications” in the same responsible chair but it was so few application and had other courses to attend that required more attention, so I couldn’t experience more in it. That’s gonna change in this semester.

Both courses are held by two very experienced professors. 3D CV is taught by Nassir Navab, whose experience in research is impressive, winning international awards and being a top recognized researcher in the world. He is going to be my lecturer in this Semester. He already began giving some introductory class showing the great advantages of 3D Computer Vision and, most of all, its applications. I was hypnotized, it was great and that’s why now I’m willing to share everything I learn and find in Internet about this matter. It’s really cool and exciting for me.

The second lecturer is Carsten Steger, who’s in charge of the class “Bildverstehen” (Image Understanding) that, as you may suppose, is in German. He is another experienced guy but his field is not in the Universities, it is the industry. He gives solution to real-life problems and that experience is shared in his class. My first impression of him is really good, he’s an open guy with a relaxed style and a very good idea of what he is talking about, besides he can truly explain himself.

Both good professors, both great courses, and an excellent semester to start experimenting with the sources of the student life. From now on you will probably see more often some posts related to Computer Vision, my new bitch (and vice versa).

But we will start with Computer Vision Online, the largest website about CV out there, I think. They offer the newest stuff about CV and its applications, besides some other curiosities and interesting sources, where you can find books, software and even you might find a suitable job for you if you are really into Computer Vision. There are tons of positions in many Universities and Labs. I recommend you to check it out!

And finally I share this video from the chair of Computer Aided Medical Procedures & Augmented Reality (CAMPAR) in the TUM, where the researchers just adapted a Kinect sensor and started to track a guy’s body to virtually visualize it as in a magic mirror. Watch it!

¡Ahí se ven!