Study On Key Technology Of Geometry Optimization For Non-planar Surface Projection System

Non-planar projection technology, which makes up the shortage of traditional planar projection technology, has a broad application prospect in the field of virtual reality, exhibition, simulation and other fields for its realistic display, flexible deployment and various expressions. There are some difficulties in the existing non-planar projection technology. First, the non-ideal optical property of the projector is seldom considered and the influence of the projector’s pose on the projection resolution and clarity is neglected. Second, the texture details of projection images are difficult to match with the shape of the object. In particular, there is no way to correct the error of projection texture when the virtual rendering model used in the generation of projection images is inconsistent with the physical screen. At last, in the present methods, the mismatch error in the overlapping projection area is corrected by manual operation which is time consuming and error prone. In view of the above problems, this paper studies the key techniques of geometric optimization in the non-planar projection system and the innovation points are:(1) Projection performance evaluation and pose optimization for projector. In order to quantitatively describe the influence of the non-ideal optical properties of the projector to the projection effect, this paper presents three kinds of projection performance evaluation criteria, including projection resolution, resolution discrepancy and projection distance discrepancy. The projection performance evaluation function is constructed by the evaluation criteria and the optimal solution of the evaluation function is obtained by multiple initial value method and Monte Carlo method. Experimental results show that the projected performance evaluation criteria can properly describe the influence of the projector pose to projection effect and the proposed method has good performance in different shape of screens. The projection resolution has been increased by 30% under the gicven optimal projector deployment which provides a useful reference for actual projector deployment.(2) Reconstruction based method for free shape surface projection. In this paper, a geometry correction method is proposed to obtain the shape of the screen by 3D reconstruction and generate the projection image by model matching and spatial coordinate transformation. In order to eliminate the possible differences between the reconstructed model and the rendering model, a texture matching error correction method based on the post image adjustment is proposed. Experimental results show that the proposed method reduces more than 60% matching error and realize the accurate alignment from the projected texture to the screen.(3) Automatic projection correction method for parametric surfaces. For surfaces which can be expressed by analytic formula, this paper proposes a non-reconstruction geometry correction method. It makes full use of the information of the parametric surface and the error caused by the manufacture or setup of the surface is considered. It achieves automatic calibration of multi-projectors in parametric surface system by using the internal constraints between the adjacent projectors. Experimental results show that the proposed method has the advantages of less time-consuming and high accuracy in the projection correction process and guarantees that the matching error is less than 0.5% of the projection system resolution.