Crop
Crop allows you to remove parts of the image that are not
of interest. Often due to NTSC conversion camera’s have spurious beginning or
ending lines that are not colored correctly. To remove those erroneous lines
from the processing pipeline use the crop function to cut those parts of the
image out. For example, if the top line of the image is bad in a 320x240 image use
X Start 0
Y Start 0
X End 320
Y End -1
Note that negative or zero numbers in the "End" coordinates will cause the current
width or height to be added before cropping to that number. This allows the crop
module to crop relative to an images size even if that size changes.
If you would like to use variables to control the crop parameters simply type in the variable
name surrounded by [] as in a [variable] expression. For example, if you enter in [image_count] into one of the provided
text box the image will be cropped according to the image count. I.e. as the image is updated image_count
is updated which will cause the cropping to change. See Set Statement to
create new variables.
Interface
Instructions
1. Dimensions - Specify the crop dimensions by either typing in the appropriate number or use the
up/down buttons to adjust the numbers. Note that numbers larger than the current
image size are ignored and do not contribute to cropping.
Note that you can use the [variable] expression format in each of the coordinate boxes in order
to programatically change the values.
2. Alignment - Given the crop coordinates specify the alignment bias of the crop. This allows you to
move the crop window relative to different sides of the image. For example, if you wanted the
right 100 pixels of the image regardless of image size you would specify xstart:0, ystart:0, xend:100, yend:0
and any of the right sided radio buttons under alignment. Note that a ending coordiante of 0 implies
the width or height of the image (it is assumed a size of 0 would not be desired).
3. Automatic - Select this checkbox to turn on automatic cropping. Images often taken as screen shots, scanned images or images from devices with extra
borders around them can be automatically cropped by detecting the box around the image in which
to crop to. The Automatic option causes the Crop module to automatically detect this box and
crop the image to those dimensions. Note that autocrop assumes that the image is surrounded by
perfectly horizontal/vertical lines having been introduced by GUI interfaces or are digitially
introduced.
4. Automatic Tolerance - When searching for the border lines you can increase the tolerance which
will allow for pixels in vertical and horizontal lines that are slightly different in color to be
considered a continious line.
5. Automatic Maximum Crop Size - Sometimes the autocrop routine will detect a much smaller box within
the image and crop too much out of the image. Specifying a low crop percentage will ensure that the
image is not cropped too much. I.e. it is assumed that the crop coordinates will be more towards the
border of the image rather than in the center.
Example
| Source | Cropped |
 |  |
Variables
CROP_X_START - holds the x start (left) of the crop transform used
CROP_Y_START - holds the y start (bottom) of the crop transform used
CROP_X_END - holds the x end (right) of the crop transform used
CROP_Y_END - holds the y end (top) of the crop transform used
See Also
Scale
|
