Difference between revisions of "Convex Mirror"
From Robert-Depot
(→V2 Background Subtract) |
|||
Line 33: | Line 33: | ||
**<code> | **<code> | ||
*accumulate error image | *accumulate error image | ||
+ | |||
+ | ==Processing from commandline== | ||
+ | *install processing-java from tools menu in processing | ||
+ | *run from command line: | ||
+ | **<code>processing-java --sketch=/Users/rtwomey/processing/dm_xml/convexmirror_cmdline --output=tmp/convexmirror --run --force</code> | ||
+ | **NOTE: all filenames in the app need full paths. |
Revision as of 07:05, 2 April 2013
Self Portrait in a Convex Mirror
Contents
Workflow
RPi side
- take photo on raspberry pi:
gphoto2 --capture-image-and-download --filename capt0004.jpg
- send to laptop:
cp capt0004.jpg rtwomey@192.168.2.1:~/Pictures/raspi
- USB reset device:
usbreset /dev/bus/usb/001/004
OS X side
- crop:
convert ~/Pictures/raspi/frame0001.jpg -crop 1450x1450+779+275 +repage square0001.jpg
- bg subtract:
python bgsubtract.py square0002.jpg square0008.jpg 20 2
- edge detect:
python edge.py square0001.jpg 20 255
- convert to pnm:
convert square0001_edge.png edge0001.pnm
- vectorize:
potrace edge0001.pnm --tight -i -s -o frame0001.svg
- send to gcode machine:
Helper script:
python process_image.py ~/Pictures/raspi/frame0024.jpg
- produces frame0024.svg
V2 Background Subtract
- subtract one image from background with given threshold and dilations:
python bgsubtract.py square0001.jpg square0002.jpg 66 3
- compute new contours from error image
- accumulate error image
Processing from commandline
- install processing-java from tools menu in processing
- run from command line:
- <code>processing-java --sketch=/Users/rtwomey/processing/dm_xml/convexmirror_cmdline --output=tmp/convexmirror --run --force
- NOTE: all filenames in the app need full paths.