Laser metal deposition is a very effective and precise technology applied by the manufacturers of turbomachinery parts to repair damages on parts in service or reclaim quality during manufacturing of new parts. The process chains to repair or refurbish turbomachinery parts are based upon a 3D laser scanning process which acquires digitalized models in the format of triangulated-meshes to represent the geometry of parts. This paper firstly studies a discrete geodesic method proposed by Mitchell, Mount and Papadimitriou (MMP) and based on this method proposes an optimized exact algorithm to generate precise laser paths on triangulated-mesh data of part geometry. Furthermore this paper presents an approximate method based on Sethian’s Fast Marching Method (FMM) and constructs vertex-based geodesic distance field on triangulated-meshes to divide the part surface in equal-distant stripes morphed from boundary curves. These two methods are compared and error condition is evaluated. During the research the methods are implemented into an automated LMD process planning software, which is utilized into industrial applications and obtained satisfactory results. The results from application of LMD process planning software are also demonstrated in this paper.