I have downloaded SDSS specra data of the dwarf galaxies for my study. I am studying strongest emission lines. I am bit confused, whether we need to perform a baseline correction beofore the measurement or not. Need your help. Thanks.
It depends on what you are measuring. If you want to measure the strength of the line, then the continuum should be subtracted, so that you have a line and a signal around zero around it (is this what you mean by baseline correction?). If you want to measure the line-to-continuum ratio or equivalent width, then you need both the measurement of the line flux and continuum flux. If you want to measure the redshift (i.e. the position of the line), then in principle this can be done on an original spectrum, but probably it will be more accurate if you subtract the continuum.
Kennicutt (1994) suggested the expression to estimate the star formation rate of the galaxies using Halpha measurement. As SFR depends on Halpha luminosity, I am trying to estimate that from SDSS spectra of the galaxy. I was confused whether I need to perform baseline correction or not before the measurement?
Yes, in order to measure the line flux (and subsequently - luminosity) you need to subtract the continuum. There is a simple explanation for this. Imagine that your galaxy has 10 times more stars but the same SFR. Then the continuum level will go up by a factor of 10, so if you don't subtract it you would measure SFR 10 times higher, which would be wrong.
By the way, check the SDSS catalogues - maybe your galaxy already has the H alpha flux measurement.