Blog

Let’s say we want to know the altitude of the drone we are flying. On one hand, our drone has a pressure sensor that allows us to measure the altitude. On the other hand, we know the altitude of the drone a second ago and the thrust produced by propellers, so the current altitude can be calculated from it. The question is how to combine these data to get one more accurate result, and often the answer is to use a Kalman filter. Kalman filter can provide an optimal estimate given a model and a measurement in the form of a normal (aka Gaussian) distribution. In this article, I will try to explain why normal distribution works and how it gives Kalman filter its superpower.

Minimalisic snake game for Vim learners: instead of using arrows you have to use h,j,k,l for navigating. If you use vimium plugin, make sure that it is disabled on this page.

Navigation: [h] left, [j] down, [k] up, [l] right

Speed: [-] slower, [0] reset speed, [+] faster

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Sed lorem ex, vulputate ut vestibulum sit amet, dignissim ut dolor. Fusce venenatis sit amet tortor sit amet porttitor. Aliquam posuere luctus lorem id finibus. Curabitur ultrices vulputate nulla mattis luctus. Nunc ligula urna, porttitor id tortor sed, consectetur ultrices ipsum. Vestibulum eleifend a libero posuere congue. Morbi non finibus quam. Etiam accumsan pharetra tellus, non iaculis nisi laoreet id. Praesent ut semper tellus. Curabitur placerat luctus turpis, id rutrum elit posuere a. Proin quam nunc, vestibulum eget elementum ac, viverra eu dolor.