Real-time Multiple Object Identification in Python

asked 2018-01-17 02:36:19 -0600

Hello all,

I have read through the OpenCV python tutorial but not gain any insights as to how real-time multiple object identification could be done - is template matching a possibility or do we have to train a cascade for it? I wish for the video feed to be overlayed with rectangular boxes around identified objects with their name underneath.

Could someone point me in the right direction, working with Python?

Thank you.

edit retag flag offensive close merge delete

Comments

Template matching will only work on very static scenes and applications. Either go for a cascade or for deep learned object detection.

StevenPuttemans gravatar imageStevenPuttemans ( 2018-01-17 04:25:15 -0600 )edit
1

please also have a look at the dnn samples

berak gravatar imageberak ( 2018-01-17 06:04:58 -0600 )edit