Hello everyone,
This is our first year using the M1101 camera. We have read the 2012 Vision whitepaper posted on CD. We are getting seemingly inconsistent errors in our distance measurements. For example, if the actual distance to the vision targets is 5ft, the measured value would be 9.2ft one time and 7.7ft another time. We have tried measuring at different distances; If we take the average of the errors across different distances it comes out to approximately 3ft.
Those of you with camera experience, can you please advise how to correct this problem?
Thanks in advance for your help.
I cannot be sure what issues are contributing, but be sure that the camera type on the Vision Processing VI and/or the Rectangular example panel is set to the M1011 and not the default of the 206. Also check the panel of the example to see if the rectangles are splotchy due to uneven lighting or other issues. Feel free to post an image showing the results.
Greg McKaskle
Greg,
A quick perusal of the vision code leads me to believe it relies on the width of the target to figure out distance. Isn’t this effected by the angle you are to the target? For instance, from one side or the other, won’t you get a misleading distance reading?
Correct. The example showed how to use a known value, and a measured value, and the view angle of the lens to estimate distance. Changing the code to use the vertical, or perhaps even multiple vertical values was left to the end-user.
It wasn’t clear that the variation mentioned above was due to view angles. Perhaps it is. Photos will tell for sure.
Greg McKaskle