A method for synthesizing an image with multi-view images includes inputting multiple images, wherein each of the reference images is corresponding to a reference viewing-angle for photographing; synthesizing an image corresponding to a viewpoint and an intended viewing-angle; segmenting the intended synthesized image to obtain a plurality of meshes and a plurality of vertices of the meshes. Each of the vertices and the viewpoint respectively establish a viewing-angle, and the method further includes searching a plurality of neighbouring images among the reference images referring to the viewing-angle. If at least one of the neighbouring images falls within an adjacent region of the vertex, a first mode is adopted without interpolation to synthesize the intended synthesized image; when none of the neighbouring images falls within the adjacent region of the vertex, a second mode is adopted, where a weighting-based interpolation mechanism is used for synthesizing the intended synthesized image.