In order to prevent point cloud noise from the surrounding environment and signal interference during the reconstruction of the architectural landscape, a visual art reconstruction method of the architectural landscape based on 3D laser printing is proposed. Kinect is used to obtain point cloud data on the surface of the object, and multi-view alignment and point cloud denoising processing are conducted to make the model data more accurate. The iterative nearest point algorithm is used to perform point cloud registration and fusion of the data, and data preprocessing is performed. The point cloud data is fused into a smooth surface model without holes. According to the three-dimensional model of the architectural landscape, the visualization system using 3D laser printer printing technology realizes the model. The simulation test results show that the error rate of the designed architectural landscape reconstruction is below 10%, and the smooth value of the reconstruction data is maintained at about 0.9, which meets the requirements of the visual art reconstruction of the architectural landscape.