Guest
Show Posts
141
Video P*r*o*b*i*n*g / Re: setting up cammera and laser
« on: December 03, 2007, 12:09:00 PM »
Hello Don,
The laser line generator is pointed down, parallel to Z axis, oriented in the YZ plane.
The camera is located a few inches to the left of the laser. It is pointed down and to the right so that it crosses the laser plane at an angle. A forty five degree angle is a good place to start experimenting.
The goal is to image the laser, and nothing else, onto the camera detector so that the laser ZY plane is imaged onto the camera HV plane. That is, the Z dimension of the laser is imaged onto the camera horizontal dimension and the laser Y dimension is imaged onto the camera vertical dimension.
This is shown at http://www.sensorland.com/HowPage056.html, except that this view shows the camera to the right of the laser.
Try to find a camera with a small angular field of view. Let's say, about 10 degrees. That will image about a 1 inch section of the laser onto the full size of the detector. If you use a wide angle camera, let's say 45 degrees, then you will probably image a 4 to 6 inch length of the laser in the Z dimension. Accuracy is some fraction, possibly 1%, of the full section imaged. So you want to image a small section to get some accuracy.
Also, short depth of field is desirable. Pinhole type webcams focus over a huge range. Let's say from 12 inches to 12 feet. You want points in the laser ZY plane to be in focus on the camera detector and you want everything else out of focus. That is, if something is closer to the camera than the laser or farther from the camera than the laser, you want that feature blurred. You cannot exactly achieve that with a prepackaged camera and lens, but you can come close. To shorten the depth of field use a larger diameter camera lens. That will collect more light from the laser while spreading background light out.
http://www.multires.caltech.edu/pubs/DGPCourse/CurlessChapter.pdf pages 4 and 5 show the correct orientation of the detector as NOT perpendicular to the lens optical axis. This is known as the Scheimpflug condition. (Say that fast three times). This is the way that the pros do it to get high accuracy. This can be done if you mount the camera separate from the lens.
Tom Hubin
thubin@earthlink.net
The laser line generator is pointed down, parallel to Z axis, oriented in the YZ plane.
The camera is located a few inches to the left of the laser. It is pointed down and to the right so that it crosses the laser plane at an angle. A forty five degree angle is a good place to start experimenting.
The goal is to image the laser, and nothing else, onto the camera detector so that the laser ZY plane is imaged onto the camera HV plane. That is, the Z dimension of the laser is imaged onto the camera horizontal dimension and the laser Y dimension is imaged onto the camera vertical dimension.
This is shown at http://www.sensorland.com/HowPage056.html, except that this view shows the camera to the right of the laser.
Try to find a camera with a small angular field of view. Let's say, about 10 degrees. That will image about a 1 inch section of the laser onto the full size of the detector. If you use a wide angle camera, let's say 45 degrees, then you will probably image a 4 to 6 inch length of the laser in the Z dimension. Accuracy is some fraction, possibly 1%, of the full section imaged. So you want to image a small section to get some accuracy.
Also, short depth of field is desirable. Pinhole type webcams focus over a huge range. Let's say from 12 inches to 12 feet. You want points in the laser ZY plane to be in focus on the camera detector and you want everything else out of focus. That is, if something is closer to the camera than the laser or farther from the camera than the laser, you want that feature blurred. You cannot exactly achieve that with a prepackaged camera and lens, but you can come close. To shorten the depth of field use a larger diameter camera lens. That will collect more light from the laser while spreading background light out.
http://www.multires.caltech.edu/pubs/DGPCourse/CurlessChapter.pdf pages 4 and 5 show the correct orientation of the detector as NOT perpendicular to the lens optical axis. This is known as the Scheimpflug condition. (Say that fast three times). This is the way that the pros do it to get high accuracy. This can be done if you mount the camera separate from the lens.
Tom Hubin
thubin@earthlink.net