search

madelonhulsebos / ComputerVision

This repository comprises the work for an academic project in the context of Computer Vision. We built a system that reads in a file, scans an image in search for a license plate, and if one was found outputs the characters on this license plate.
6 stars 5 forks source link

Viola & jones implementation for license plate detection #10

Closed madelonhulsebos closed 7 years ago

madelonhulsebos commented 7 years ago

Implemented the Viola & Jones approach for detecting license plates in images according to their paper[1]. The model License plate detection/Viola & Jones/classifiers/vj bdt* trained on the datasets/acme licenses/centered opaque rotated background dataset does not perform well on the test images in the testset datasets/markus cars: resulting in too many false positives (~30 per image). This approach therefore will not be used to detect license plates in the final system.

[1] Viola, P., & Jones, M. (2001). Rapid object detection using a boosted cascade of simple features. In Computer Vision and Pattern Recognition, 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on (Vol. 1, pp. I-I). IEEE.

JeroenMandersloot commented 7 years ago

LGTM

One question: is there a reason there are two copies of the classifiers folder and the Viola_Jones_train.py file? One copy is directly below the License plate detection directory, and another copy is in License plate detection/Viola & Jones.

madelonhulsebos commented 7 years ago

I forgot to commit those changes, thanks for noticing!

JeroenMandersloot commented 7 years ago

Approved!