Hello everybody,i am traying to make robot which must built a map. Robot will built a map.the map is 2d.the mapping is into the indoor..
i will use ultrasonic sensor.the sensor measure the rangeof indoor.How can i do the built mapping. The robot goes and measure range of the room and the ranges will be in a array.and i will use bluetooth module with communication.
And my teacher says to me. You will be matlab gui for mapping.but i didn't how can i built a map with the array which is measure from sensor.
İs there any people research with Slam.Can you help me.
Finally i am sorry for my bad english.i know that.
thanks a lot
How can i do the built mapping.
Your best bet is to do extensive searching using Google to find similar projects. The ultrasonic sensor may not be good for generating any type of useable map.
aforizma90:
Hello everybody,i am traying to make robot which must built a map. Robot will built a map.the map is 2d.the mapping is into the indoor..
Ok - you're doing a 2D indoor mapping robot; got it.
aforizma90:
i will use ultrasonic sensor.the sensor measure the rangeof indoor.How can i do the built mapping. The robot goes and measure range of the room and the ranges will be in a array.and i will use bluetooth module with communication.
Ok - you're going to use an ultrasonic sensor to detect the ranges in 2D, and store those ranges into an array; what will this array be representing? Will it represent a 2D map of obstacle edges? Is it a 1D array of vertex points representing edges of obstacles? Lot's of possibilities here...
As far as the bluetooth module is concerned - are you planning on doing all mapping and calculation on a PC? Please note that if you are planning on doing SLAM (as you note later), this will be the only likely way to do it (although I have given thought to the idea of SLAM on an expanded Mega - in theory it is possible).
aforizma90:
And my teacher says to me. You will be matlab gui for mapping.but i didn't how can i built a map with the array which is measure from sensor.
I'm not very familiar with MatLab; the closest thing I have used to MatLab in the past is the open-source clone called Octave (it doesn't have a GUI in its standard form, it's a command-line system - though IIRC there are GUI tools for it).
It's been too long since I played with it, though - and when I did, I used GNUplot for graphing output...
aforizma90:
İs there any people research with Slam.Can you help me.
My only experience with SLAM was taking the Udacity CS373 MOOC:
https://www.udacity.com/course/cs373
If you have the time, I strongly suggest you taking it - it is based around Python, though - not MatLab or Octave. You will need to have a good understanding of linear algebra, and more importantly, probabilities/statistics - this is a basic requirement of just about all AI/ML anyhow. I can say though that I found the course to be very enlightening, and that it explained a lot about SLAM, and other topics that I (and others) struggle with understanding (particularly Kalman filtering and PID).
Beyond that, take a look at the following resources:
Pay special attention to the external resource link regarding "SLAM for Dummies":
This "tutorial" link is actually only one PDF in what I found to be multiple - you'll have to do some googling, but there is more information out there about it (some further links might be found within that document). A good set of search terms for SLAM on Google are:
"slam simultaneous localization and mapping"
Here is a resource of various open-source SLAM algorithms; note, though, that if you don't understand the basics of SLAM (and even if you do!) - you likely won't understand these implementations very much:
Not every SLAM algorithm is open-source or non-proprietary; SLAM is not a "solved" problem by any means. Multiple volumes have been written about it, tons of research papers are available (I encourage you to seek them out). It is one of those research areas which is "hot" - lots of work is being done in the area, and if a particular general solution could be found (unlikely as it may be), it could mean a lot of money to the person or group behind it - which is why some of the solutions are considered proprietary and non-published.
aforizma90:
Finally i am sorry for my bad english.i know that.
thanks a lot
Much better than my foreign language skills...! 