Have a look at the car detector. As it seems your spot are well defined (you could have static bounding box for each car spot), you have to detect if a car is present or not.
The idea is to train a Cascade Classifier with HAAR or LBP, or a HOG Detector. You could find some nice tutorial online for that, as well as book. The training code is provided by OpenCV also here.
As the spots are well defined and assuming camera is fixed, you could use opencv background subtraction with sufficient amount of history to detect if there is an object in a particular parking spot. Look at this opencv tutorial for more information. http://docs.opencv.org/master/d1/dc5/...
Asked: 2016-02-23 05:05:24 -0600
Seen: 2,478 times
Last updated: Feb 23 '16
