ptraykovski wrote:Seems like vReveal is gone:
ptraykovski wrote:About a year ago I tried some horizon detection algorithms and when a clear horizon was visible they worked great, but failed drastically when it was not visible. I might revisit some of the image based techniques in combo with IMU. This might help provide a initial point of search for the image based techniques.
Using the IMU I noticed on my footage that kite mounted corrections had some problem when the kite was accelerating quickly, but string mounted seems better as much lower accelerations. So was testing methods to fix this issue over the weekend with two IMUs mounted on a windsurfing mast. One near the center of rotation and the other near the end of the mast as a swung it around.
As far as selling this goes, I would probably have to find a lower cost sensor as the one i am using now is ~ 400$. There are some around 100$ but have no idea how good they are. It will probably come down soon with all the wearable tech etc. The IMU chips alone are cheap but it is a lot of work to go from a chip to a fully calibrated sensor with data fusion algorithms that produces the output I need. I don't want to repeat that. Also the matlab environment I am doing the software in isn't ideal for sharing except with those who have licences. But I will try to find so some workaroundsfor these issues in the future.
What would people be willing to pay for some thing like this?
Users browsing this forum: Bing [Bot] and 16 guests