We present a study of low surface brightness galaxies(LSBGs) selected by fitting the images for all the galaxies inα.40 SDSS DR7 sample with two kinds of single-component models and two kinds of two-component models(disk+bulge):single exponential,single sersic,exponential+deVaucular(exp+deV),and exponential+sérsic(exp+ser).Under the criteria of the B band disk central surface brightness μ0,disk(B)≥22.5 mag arcsec-2and the axis ratio b/a> 0.3,we selected four none-edge-on LSBG samples from each of the models which contain 1105,1038,207,and 75 galaxies,respectively.There are 756 galaxies in common between LSBGs selected by exponential and sersic models,corresponding to 68.42% of LSBGs selected by the exponential model and 72.83% of LSBGs selected by the sersic model,the rest of the discrepancy is due to the difference in obtaining μ0between the exponential and sersic models.Based on the fitting,in the range of 0.5≤n≤1.5,the relation of μ0from two models can be written as μ0,sérsic-μ0,exp=-1.34(n-1).The LSBGs selected by disk+bulge models(LSBG2comps) are more massive than LSBGs selected by single-component models(LSBG1comp),and also show a larger disk component.Though the bulges in the majority of our LSBG2comps are not prominent,more than 60% of our LSBG2comps will not be selected if we adopt a single-component model only.We also identified 31 giant low surface brightness galaxies(gLSBGs) from LSBG2comps.They are located at the same region in the color-magnitude diagram as other gLSBGs.After we compared different criteria of gLSBGs selection,we find that for gas-rich LSBGs,M*> 1010M⊙is the best to distinguish between gLSBGs and normal LSBGs with bulge.