Easy home-position-calibration with a web-cam
Posted by VDX
|
Easy home-position-calibration with a web-cam December 23, 2007 03:19PM |
Hi all,
... because i had some displacement of the dispenser-tip every time i changed the dispenser, i tried to define or reset the new home- or Null-position with a web-cam.
Here the images of a 0,2mm- and a 0,9mm-dispenser-needle, viewed with a self-made micro-cam:
With the camera embeded upside-down in the corner where the X- and Y-axes have the home-switches i can easily define a very exact Null-position for my dispenser, when after homing manually moving the toolhead with the dispenser in a specific position in relation to the cam-image - best marked with a cross-hair on a plastic-sheet on the screen.
So i have an absolute exact start-position every time i change the dispenser -- the camera can be optimized for 3 to 5 micron pixel-size, so i would have an Null-position-accuracy of nearly 0,02 to 0,01 millimeter, what's better then homing with a mechanical switch and redefine by guess ...
The camera can be a cheap web-cam: - when you extract the lens-head, turn with the inner lens outside and adapt a black tube, so that the lens-head is some centimeters from the camera-chip, you convert the macro-webcam in a high-resolution-video-microscope!
With a bit more optimizing i built microscope-cameras with resolutions beneath 1 micron and a field of view of 0,8x0,5mm
But for the accurate homing i need 'only' an area of 8x5 to 5x3 mm, so you can play with the tube-length ...
Viktor
... because i had some displacement of the dispenser-tip every time i changed the dispenser, i tried to define or reset the new home- or Null-position with a web-cam.
Here the images of a 0,2mm- and a 0,9mm-dispenser-needle, viewed with a self-made micro-cam:
With the camera embeded upside-down in the corner where the X- and Y-axes have the home-switches i can easily define a very exact Null-position for my dispenser, when after homing manually moving the toolhead with the dispenser in a specific position in relation to the cam-image - best marked with a cross-hair on a plastic-sheet on the screen.
So i have an absolute exact start-position every time i change the dispenser -- the camera can be optimized for 3 to 5 micron pixel-size, so i would have an Null-position-accuracy of nearly 0,02 to 0,01 millimeter, what's better then homing with a mechanical switch and redefine by guess ...
The camera can be a cheap web-cam: - when you extract the lens-head, turn with the inner lens outside and adapt a black tube, so that the lens-head is some centimeters from the camera-chip, you convert the macro-webcam in a high-resolution-video-microscope!
With a bit more optimizing i built microscope-cameras with resolutions beneath 1 micron and a field of view of 0,8x0,5mm
But for the accurate homing i need 'only' an area of 8x5 to 5x3 mm, so you can play with the tube-length ...
Viktor
|
Re: Easy home-position-calibration with a web-cam December 23, 2007 04:21PM |
Hi Jay,
... here a simple sketch:
The resolution is higher, when the turned optic is far away, but then the image is darker too.
Best lighting is with super-bright white-LED's, as the normal cameras are more sensitive in the blue range - a possible hack is to break away the IR-filter mostly glued on top of the glass-window of the camera-chip.
When the light is good (or you can rise the sensitivity of the chip), then it's possible too to implement a blind with a small aperture - so i can rise the depth of focus until a decade, when balancing with light and camera-sensitivity.
Here you can see a comparison of images made with 'normal' video-optics and some i optimized for better field of view:
[mikroskopkamera.de]
Viktor
... here a simple sketch:
The resolution is higher, when the turned optic is far away, but then the image is darker too.
Best lighting is with super-bright white-LED's, as the normal cameras are more sensitive in the blue range - a possible hack is to break away the IR-filter mostly glued on top of the glass-window of the camera-chip.
When the light is good (or you can rise the sensitivity of the chip), then it's possible too to implement a blind with a small aperture - so i can rise the depth of focus until a decade, when balancing with light and camera-sensitivity.
Here you can see a comparison of images made with 'normal' video-optics and some i optimized for better field of view:
[mikroskopkamera.de]
Viktor
Sorry, only registered users may post in this forum.