Dear all,
I am using ADXL335 accelerometer for the purpose of detecting the hand motion. I am not able to derive the exact relation between the accelerometer values with cartesian co-ordinates. Please help to map the accelerometer data to cartesian co-ordinate points.
Thank you.
Dear Nishith Kotak
It is not neсessary or impossible.
Coordinate system of ADXL335 is cartesian: sensitive axis of uniaxial accelerometers are orthogonal . Accelerometer data are projection to cartesian coordinate of 3D sensor. But this is mobile coordinate system.
Perhaps you mean the projection of accelerations on the fixed axis associated with the floor. This projections you get will not be able without additional sensors that can determine the orientation of the axes of the accelerometer. Typically used angular velocity sensors (accelerometer together they are the inertial unit), motion capture system or others. This problem is complicated, and I recommend to look for a assembly if you check the accuracy of execution.
Also You can take into account: accelerometer measurement is vector sum of sensor body acceleration and gravitational acceleration.
It is possible use pair of accelerometers and theorems of mechanics.
Regards
Pavel Kruchinin
Hello Nishith Kotak,
as Pavel Kruchinin pointed already, the accelerometer gives its output in cartesian coordinate, but your problem is that you wanted those coordinates to be on a global coordinate frame and not in the body frame. For you to solve your problem, you would need an additional sensor which is expressed with respect to a global frame -like the magnetometer, for instance. If you had a magnetometer tightly attached to your accelerometer, you could find the angle difference between the gravity vector and the vector of the magnetic field estimated with the magnetometer, and with this angle rotate your acceleration into a proper "global" frame.
I attach a publication where what I mentioned before about changing frames is explained. Although the attached paper the sensors are attached to the foot, you could as well get some inspiration about pose detection for hands ;)
I hope I helped you.
Best regards,
Daniel Medina
Conference Paper A study on indoor pedestrian localization algorithms with fo...
- Badges
- Science topic
- Similar topics
- Condensed Matter Physics
- Zero