Hardware manual

Camera Impact Reference Guide
Datalogic Automation Inc. 2-30
2. Enter the Units per Pixel value in the Manual Settings group. Units per Pixel is the scale factor for con-
verting measurements in the image to "real-world" units. For example, if an object measures 100 milli-
meters and the image resolution of that object is 640 pixels, the Units per Pixel value is 100 divided by
640 = 0.156 (or 0.156 mm per pixel).
3. Choose either Disable Lens Distortion Calibration or Keep Current Lens Distortion Calibration. This
selection determines whether the distortion information is stored with the image or not.
4. Click the Calibrate button.
Manual Settings
• Units per Pixel
This is the factor that VPM tools use to convert image measurements to "real-life" measurements. For
example, if you want to measure a part that is 100 mm wide and the image window resolution is 640
pixels wide, each pixel is equivalent to 0.156 mm:
100 mm / 640 pixels = 0.156 mm per pixel
• Disable/Keep Lens Distortion Calibration
Most lenses introduce some distortion in images. When a camera is calibrated, the calibration process
can account for most of the natural pixel size and perspective distortion (not radial distortion) present
in the lens. You can choose whether to disable or preserve that lens distortion in each image taken by
the camera.
Calibrate From Polygon Points
Use this option when it is difficult to calibrate with a target or when the inspected image contains perspec-
tive distortion and you can align the provided polygon over points on the image. This option will correct for
pixel size and perspective distortion, but not radial distortion. To calibrate for radial distortion you must use
a calibration target.
The polygon points and real-world points must correspond one-to-one, meaning that there must be the same
number and they must be in the same order. The polygon points are ordered clockwise. To see a tooltip
showing the point order and X/Y coordinates, hold the cursor over the desired point. Once you know the
order of the polygon points, you can enter the real-world points in the correct order.
To calibrate for real-world origin and pixel size calibration, the points you select should represent at least
three distinct X and Y locations on the part. For acceptable distortion correction, select at least four distinct
X and Y locations on the part. For example, if you only use points along the top and bottom edges, the dis-
tortion correction in the vertical direction is ill-defined and will likely be unacceptable because only two dis-
tinct Y locations are represented. You should try to space the points into at least four distinct X and Y
locations.
The calibration algorithm will automatically determine whether to calibrate for distortion or not. If there is
not detectable distortion, or if there are too few points, it will calibrate for real-world origin and pixel size. If
there is detectable distortion and enough points, it will also calibrate for distortion. The current setting is dis-
played after the Lens Distortion label in the “Current Calibration” box below the image window.
It is possible to “confuse” the distortion calibration algorithm by defining certain point positions. If you get
“odd” calibration results, you should first check that you entered the correct real-world coordinates in the
correct order. If these are correct, use more points to better define the distortion.
NOTE: Since the image origin (0,0) is in the upper-left corner of the image, the X and Y real-world coordi-
nate values use a "left-handed" coordinate system. This means that the X coordinate value increases to the
right and the Y coordinate value increases downward. See “Point Settings” below.