Requirements to use OpenCV with
Scorpion 8
To use OpenCV in Scorpion Vision you must
follow the Python 2.6
installation
guide.
OpenCV 2.2 setup
There are some additional requirements for OpenCV 2.2
- Install numpy (1.5.1) from http://new.scipy.org/download.html
- Install OpenCV 2.2
- Make sure that OpenCV "bin" folder is added to system environment path variable (it can be selected when installing
or added later via advanced system settings).
Default installation would be "C:\OpenCV2.2\bin"
- Download Python 2.6 wrapper for OpenCV 2.2 here
- Unzip python 2.6 wrapper for OpenCV 2.2 into OpenCV installation folder (or any other
accessible location)
OcvImage - Scorpion helper class to
use OpenCV
The OcvImage python module is located
here for OpenCV 2.1 and 2.2.
Sample profiles with OpenCV 2.2
Note: For OpenCV 2.1 and 2.2 use import cv
Note2:
- For OpenCV 2.2 samples paths to OpenCV library are set to fixed location
- C:\OpenCV2.2\Python2.6\Lib\site-packages.
- Make sure it points
to correct location on your machine.
Example 1: ScriptTool
cvAdaptiveThreshold
from OcvImage21 import OcvImage
import cv
objectImage = OcvImage(GetTool(GetCurSource()).imageNo)
object = objectImage.GetCvImage()
try:
#Example of a useful transform - adaptive threshold
#AdaptiveThreshold(im,im2,max_value,adaptive_method,threshold_type,block_size,offset)
#apply to OpenCV image, can be done in place
cv.AdaptiveThreshold(object,object,125,cv.CV_ADAPTIVE_THRESH_MEAN_C,cv.CV_THRESH_BINARY_INV,7,10)
#can apply morphology if desired,
# - again safe to do in place or could make second image
cv.Dilate(object,object,None,1)
# cv.Erode(object,object,None,2)
OcvImage().Assign(object, "Output")
except Exception, msg:
print msgExample
2: Adding Canny Filter
#the script reside inside a Scorpion
ScriptToolfrom OcvImage21 import OcvImage
import cv
objectImage = OcvImage(GetTool(GetCurSource()).imageNo)
object = objectImage.GetCvImage()
try:
cv.Canny(object, object,40, 60,3)
OcvImage().Assign(object, "Output")
except Exception, msg:
print msgInternet Resources on OpenCV with
Python
|