We are about to switch to a new forum software. Until then we have removed the registration on this forum.
Good day everybody,
for disclaimer: yes I am going to ask for help with a university project, but I won't ask you to do the work for me. Just push me in the right direction, please
So here is the deal. I'm supposed to write an advent calendar in processing. It should film the reality, find a door and open this door virtually with something christmassy behind.
I have trouble to think of a way to detect a door. I've looked into the OpenCV library and the canny edge detector looks promising. I could use the canny function on the filmed image and then somehow compare its edges with the edges of an example image of an actual door. This is where I am stuck. And I'm not sure what I could google that could help me. Maybe you have an idea...
(In case this idea is not a good practice, please recommend other libraries or ways.)
I haven't started yet so I don't have any code-examples. I'm just trying to figure out the concept. Btw I'm an absolute newbie to processing so please be patient
I'm using Processing 2.2.1 and OpenCV for Processing 0.5.0 by Greg Borenstein
Answers
I haven't used OpenCV myself, but try having a look at the documentation. The port for Java/Processing probably has similar documentation. There is an object detection module, which might fit your needs:
http://docs.opencv.org/modules/ocl/doc/object_detection.html
Specifically I'm thinking of the matchTemplate functions, but there is a "cascade classifier" as well there, that could be used.
Or for something similar that doesn't require OpenCL as well: http://docs.opencv.org/modules/imgproc/doc/object_detection.html#void matchTemplate(InputArray image, InputArray templ, OutputArray result, int method)
Thanks! Thats sounds good. I will look into this!