If this is solely vision based then this is stereo vision right? Or are you using some form of rangefinder?
Also, you say the blue line is the uncorrected robot odometer reading, how is this corrected? Via the map it builds on the fly? or with a compass, or is this corrected afterwards for this display?
I find this very interesting, please share some more detail! Thanks :)
How you've managed to export the data from the sensor(camera) to a visual representation in the computer? Did you use, lets say matlab, or another software package?If yes, which one? Thanks in advance.
If anyone else is looking for more info on this, Google 'SLAM vision' for a heap of research papers etc. Really interesting stuff.
mrquan83 4 years ago
If this is solely vision based then this is stereo vision right? Or are you using some form of rangefinder?
Also, you say the blue line is the uncorrected robot odometer reading, how is this corrected? Via the map it builds on the fly? or with a compass, or is this corrected afterwards for this display?
I find this very interesting, please share some more detail! Thanks :)
mrquan83 4 years ago
This is done using stereo vision with SIFT features and particle filters. The occupancy grid is generated using the correlation-based stereo images.
awesomo2001 4 years ago
Which simulation program did you use?
and what is the blue line?
nospa4000 4 years ago
This is not simulation; the grid is constructed from real data. The blue line is the uncorrected robot odometry.
awesomo2001 4 years ago
How you've managed to export the data from the sensor(camera) to a visual representation in the computer? Did you use, lets say matlab, or another software package?If yes, which one? Thanks in advance.
nospa4000 4 years ago