rectangular object detection in video stream using the Generalized Hough Transform

This clip shows the result of applying the Generalized Hough Transform in image frames grabbed from a web-camera in order to detect rectangular objects in the frames originating from the video stream. In the video, the detected objects are captured in green closed lines, indicating the contour of the object.
Hough transform can be reformulated in order to detect any kind of object.

Category:

Education

Tags:

• hough transform
• opencv

[zzzROSSzzz]

have you got some answers, dude?...

what are you using there? opencv? if yes on which OS (Windows or Linux)?

i'd be pleased to get any recommendations, coz i need to do object detection for a university project

thanks in advance!!

[brigr]

@zzzROSSzzz It's OpenCV on Ubuntu Linux... Go to Google scholar; you'll get enough articles from there on the subject... Also IEEEexplore would be helpful if you have access to it...

[ricardopcunha]

Is your example open source?

[brigr]

@ricardopcunha No it was just some experimental code that I worked on in the lab a year ago... Sad news is that I do not have the code any more... I used a few papers from Google Scholar in order to work out the original version of the hough transform in opencv...

[pnitsch]

Brilliant. This uses OpenCV?

[brigr]

@pnitsch Yes, it used a slight variant of HT.

[brigr]

@ricardopcunha No it was not open sourced; it was just an experiment i did several days in the lab...

[brigr]

@hediondon This comment is realistic, judging from the fact that I am in direct contact with the problem each day. However, at least I feel I could contribute something to the community in the framework of this research. :)

[hediondon]

Nice work. lol at your research you must be sick of porn lol.