A person can be uniquely identified by evaluating one or more distinguishing biological traits. One such trait is scanned images of hand writing. It is useful biometric modality with application in forensic analysis, banking sector, document verification etc. This study focus on text-independent writer identification method based on scale invariant feature transform (SIFT). It consists of three phase: training, enrolment and identification phase. In all of these three phases an isotropic LoG filter is applied first. It is used to segment the handwriting image into word regions (WRs) and then, the SIFT descriptors (SDs) of word region and the corresponding scales and orientations (SOs) are extracted. In the first phase, an SD codebook is constructed by gathering the SDs training samples. In the second phase, SD signatures (SDS) are generated by looking up the SD codebook and SOs is used to generate a scale and orientation histogram (SOH). And finally in last phase, the SDS and SOH of the input handwriting are extracted and matched with the enrolled ones for identification. Here instead of analysing handwritten document we experiment on signature of an individual. Using SIFT algorithm we demonstrate that he/she is genuine or not.