Signature　verification　is　an　important　part　of　digital　forensics.　In　order　to　solve　the　shortcomings　of　manual　identification　in　technical　accuracy　and　subjectivity,　this　paper　proposed　an　off-line　signature　identification　method　based　on　Support　Vector　Machine　(SVM).　A　powerful　feature　set　is　collected　by　extracting　grid　features　and　global　features　of　a　signature　picture.　The　method　is　applied　for　identifying　different　writing　systems　and　the　highest　correct　probability　of　identification　arrives　at　100%.　The　results　indicated　that　the　method　is　workable　and　can　be　an　effectively　technical　support　for　digital　forensics.　©　2016　IEEE.