Its depend what kind of analysis you are look. If you want to do through RS data (two or three years), simply classify ( Unsupervised, Supervised, OBIA, etc.). ERDAS, ENVI, etc. can be use.

If you have bulk of data then Timesat is best

Yes, i agree. You can make classification of vegetation with erdas imagine (remote sensing), but you should have land cover elements to classify that with accuracy.

Have you considered a micro UAV or UAS with a NDVI cameras or IR and UV cameras. The UAV use GPS so it is very easy to link the data gathered with GIS. The systems a not the expensive and you can target individual fields/crops.

There are different models rotor models can hover better for single fields or live streaming compared to winged models that's are better over large areas.

Have a look at

http://www.questuav.co.uk or http://www.draganfly.com

Thank for yours valuable comments thanks sir

I think NDVI is the standard method used for vegetation detection. Classification is used when you are about to detect the change occurred during a span of time. Be careful when using classification because you need to calibrate your data before conducting the classification.