For　target　recognition　of　sparse　LiDAR　scan　data,　this　paper　proposes　a　new　method　for　aligning　height　data　matrices　based　on　rotation,　length　segmentation,　and　grid　partitioning.　Initially,　the　method　involves　rotating　the　matrix　to　achieve　initial　pose　alignment　between　the　height　matrix　of　the　target　to　be　recognized　and　the　template　matrix.　Subsequently,　utilizing　length　segmentation,　the　orientation　of　the　target　is　aligned,　and　the　initial　grid　search　position　is　determined.　Finally,　the　method　searches　for　the　best　alignment　grid　based　on　grid　partitioning　to　achieve　data　matrix　alignment　and　performs　target　recognition　using　a　fuzzy　recognition　algorithm　based　on　scanlines.　Experimental　results　demonstrate　that　this　method　effectively　aligns　sparse　target　data　with　target　template　data,　also　enabling　the　recognition　of　typical　targets　such　as　cars,　providing　a　valuable　reference　for　target　recognition　of　sparse　LiDAR　scan　data.　©　2024　SPIE.