The determination of N (to be converted into protein by some factor) and hexane soluble material (called lipids) is quite straightforward. Starch was and will probably always remain some what more tricky. Conversion into monomers (mostly sugars) has to be carried out in some way and then there are quite some methods described to determine the sugar content.
It sounds not to be easy at all if I would like to measure the accurate component of rice seed, we do not have enough the power source. I may take simple way to do this for proximate analysis only.
Proximate analysis would be the appropriate way to analyse. Once you make your samples in a form of a powder with dry matter you can analyse crude protein (by multiplying by the relevant factor and crude fat and also starch content by removing the soluble sugars and converting starch into reducing sugars. Analyse the reducing sugars by a simple method such as Lane & Eynone reducing sugar analysis and by multiplying by .9 and can find the total starch.