This paper presents a method to reconstruct symmetric geometric models from point cloud with inherent symmetric structure. Symmetry types commonly found in engineering parts, i.e., translational, reflectional and rota...This paper presents a method to reconstruct symmetric geometric models from point cloud with inherent symmetric structure. Symmetry types commonly found in engineering parts, i.e., translational, reflectional and rotational symmetries are considered. The reconstruction problem is formulated as a constrained optimization, where the objective function is the sum of squared distances of points to the model, and constraints are enforced to keep geometric relationships in the model. First, the explicit representations of symmetric models are presented. Then, by using the concept of parameterized points (where the coor-dinate components are represented as functions rather than constants), the distances of points to symmetric models are deduced. With these distance functions, symmetry information, for both 2D and 3D models, is uniformly represented in the process of reconstruction. The constrained optimization problem is solved by a standard nonlinear optimization method. Owing to the explicit representation of symmetry information, the computational complexity of our method is reduced greatly. Finally, examples are given to demonstrate the application of the proposed method.展开更多
基金the National Natural Science Foundation of China (No. 50575098)China Postdoctoral Science Foundation (No. 20070421176)
文摘This paper presents a method to reconstruct symmetric geometric models from point cloud with inherent symmetric structure. Symmetry types commonly found in engineering parts, i.e., translational, reflectional and rotational symmetries are considered. The reconstruction problem is formulated as a constrained optimization, where the objective function is the sum of squared distances of points to the model, and constraints are enforced to keep geometric relationships in the model. First, the explicit representations of symmetric models are presented. Then, by using the concept of parameterized points (where the coor-dinate components are represented as functions rather than constants), the distances of points to symmetric models are deduced. With these distance functions, symmetry information, for both 2D and 3D models, is uniformly represented in the process of reconstruction. The constrained optimization problem is solved by a standard nonlinear optimization method. Owing to the explicit representation of symmetry information, the computational complexity of our method is reduced greatly. Finally, examples are given to demonstrate the application of the proposed method.
