idea about difference bet moving red obj & fire [closed]

asked 2017-04-18 07:05:11 -0500

sabra gravatar image

the subject is clear friends, i can detect fire very good by svm, but when i add negative sample about moving red obj, again in the test step , a moving red object detected as fire!!

edit retag flag offensive reopen merge delete

Closed for the following reason the question is answered, right answer was accepted by sturkmen
close date 2020-10-15 14:16:33.800610


Do you have many moving red objects in your training data? You should make sure your training data has examples of all the types of things you want to be able to separate.

Tetragramm gravatar imageTetragramm ( 2017-04-18 17:44:09 -0500 )edit

Without your complete pipeline and some examples, it is kind of impossible to say why they get detected. My best guess based on your previous posts, is that you are using HSV color based pixel classification. Probably your ranges are too close to the HSV values for moving red objects. As long as you stick to color only, you will never solve this issue.

StevenPuttemans gravatar imageStevenPuttemans ( 2017-04-19 05:04:04 -0500 )edit

hsv is better or rgb or ycrcb? im using: colorDetect(RGB based)+svm.predict

if i add red moving obj to negative,true alarms reduced


sabra gravatar imagesabra ( 2017-04-21 23:33:04 -0500 )edit

in my opinion, HSV seperates colors better than RGB, but still color based segmentation will only get you so far ...

StevenPuttemans gravatar imageStevenPuttemans ( 2017-04-24 06:03:34 -0500 )edit