How to combine two or more extracted features from a video in a single variable?

Hi everyone...

I'm doing my project on abnormal behaviour detection in video.

First I have applied static and dynamic agents, then optical flow algorithm on the input video...

Then i have extracted features such as BRISKFeatures, FASTFeatures, HarrisFeatures, MinEigenFeatures, SURFFeatures.

My question is, will it possible to combine all these features and then train it? Is it a correct way? If not then what is the way to train with all these features?

Please guide me to get through....

I could be very Thankful...

Thank you...