what is the best methord for processing this type of image ie,identifing segments

segmentation

asked 2016-08-02 13:28:33 -0600

shek
1

sensor output

i want to identify objects in this type of images . this is just an example of the sensor output

edit retag flag offensive close merge delete

Comments

Start off with known pedestrian detectors like Viola&Jones, HOG+SVM, LatentSVM, ICF/ACF, ... if you wanna tackle person detection!

StevenPuttemans ( 2016-08-03 04:25:38 -0600 )edit

add a comment

1 answer

Sort by » oldest newest most voted

answered 2016-08-02 19:11:06 -0600

mrquorr

What do you want to identify? Will it always have the same rotation and general form? For example, identifying the men will need a different method if they could be on different angles of rotation or with different type of clothes, or if you need it to be quickly identified, or with a certain percentage of certainty.