This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
a binary classifier (such as theCascadeClassifier) can only detect presence or absence of 1 thing.
you can train it to find an open door (vs. the rest of the world) or you can train it to detect a closed door(again, vs. the rest of the world).
you cannot train it to a) find a door, and b) then distinguish, if it's open or closed.
| | 2 | No.2 Revision |
a binary classifier (such as theCascadeClassifier) can only detect presence or absence of 1 thing. train a
unfortunately, you can train it to find an open door (vs. the rest of the world) or you can train it to detect a closed door(again, vs. the rest of the world).have 3 classes already here, door-open, door-closed and not-a-door.
you cannot a single binary classifier won't be able to solve this.
probably the best idea is a 2 stage approach, first train it a cascade classifier to a) find detect doors in general (both open/closed), against all other parts of a door, and b) then distinguish, if it's house, then,
train another classifier (maybe an SVM) to distinguish between open or closed.closed door from the detected region .
| | 3 | No.3 Revision |
a binary classifier (such as theCascadeClassifier) can only detect presence or absence of 1 thing. train a
unfortunately, you have 3 classes already here, door-open, door-closed and not-a-door.
a single binary classifier won't be able to solve this.
probably the best idea is a 2 stage approach, first train a cascade classifier to detect doors in general (both open/closed), against all other parts of a house, then,
train another classifier (maybe an SVM) to distinguish between open or closed door from the detected region .
