Go-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registration This paper presents Go-ICP, the first globally optimal algorithm for 3D point-set registration under the L2 error metric. Go-ICP is based on a branch-and-bound (BnB) scheme that searches the entire 3D motion space SE(3). By exploiting the special structure of SE(3) geometry, novel upper and lower bounds for the registration error function are derived. Local ICP is integrated into the BnB scheme, which speeds up the new method while guaranteeing global optimality. The method is robust to outliers and can be applied in scenarios where an optimal solution is desirable or where a good initialization is not always available. The paper discusses the problem of 3D point-set registration, which is a fundamental problem in computer and robot vision. The goal is to find the transformation that best aligns one of the point-sets to the other. The Iterative Closest Point (ICP) algorithm is the most well-known algorithm for efficiently registering two 2D or 3D point-sets under Euclidean (rigid) transformation. However, ICP is susceptible to local minima due to the non-convexity of the problem and the local iterative procedure it adopts. To address this issue, previous efforts have focused on widening the basin of convergence, performing heuristic and non-deterministic global search, and utilizing other methods for coarse initial alignment. Go-ICP is based on the well-established Branch-and-Bound (BnB) theory for global optimization. The method is named the Globally Optimal ICP, abbreviated to Go-ICP. The proposed method always produces the exact and globally optimal solution, up to the desired accuracy. The algorithmic structure of the proposed method can be summarized as follows: use BnB to search the space of SE(3), whenever a better solution is found, call ICP initialized at this solution to refine the objective function value. Use ICP's result as an updated upper bound to continue the BnB. Until convergence. The error metric strictly follows that of the original ICP algorithm, minimizing the L2 norm of the closest-point residual vector. The method is also extended with robust kernels or robust norms. A preliminary version of this work was presented as a conference paper. The paper discusses the problem of 3D point-set registration, which is a fundamental problem in computer and robot vision. The goal is to find the transformation that best aligns one of the point-sets to the other. The Iterative Closest Point (ICP) algorithm is the most well-known algorithm for efficiently registering two 2D or 3D point-sets under Euclidean (rigid) transformation. However, ICP is susceptible to local minima due to the non-convexity of the problem and the local iterative procedureGo-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registration This paper presents Go-ICP, the first globally optimal algorithm for 3D point-set registration under the L2 error metric. Go-ICP is based on a branch-and-bound (BnB) scheme that searches the entire 3D motion space SE(3). By exploiting the special structure of SE(3) geometry, novel upper and lower bounds for the registration error function are derived. Local ICP is integrated into the BnB scheme, which speeds up the new method while guaranteeing global optimality. The method is robust to outliers and can be applied in scenarios where an optimal solution is desirable or where a good initialization is not always available. The paper discusses the problem of 3D point-set registration, which is a fundamental problem in computer and robot vision. The goal is to find the transformation that best aligns one of the point-sets to the other. The Iterative Closest Point (ICP) algorithm is the most well-known algorithm for efficiently registering two 2D or 3D point-sets under Euclidean (rigid) transformation. However, ICP is susceptible to local minima due to the non-convexity of the problem and the local iterative procedure it adopts. To address this issue, previous efforts have focused on widening the basin of convergence, performing heuristic and non-deterministic global search, and utilizing other methods for coarse initial alignment. Go-ICP is based on the well-established Branch-and-Bound (BnB) theory for global optimization. The method is named the Globally Optimal ICP, abbreviated to Go-ICP. The proposed method always produces the exact and globally optimal solution, up to the desired accuracy. The algorithmic structure of the proposed method can be summarized as follows: use BnB to search the space of SE(3), whenever a better solution is found, call ICP initialized at this solution to refine the objective function value. Use ICP's result as an updated upper bound to continue the BnB. Until convergence. The error metric strictly follows that of the original ICP algorithm, minimizing the L2 norm of the closest-point residual vector. The method is also extended with robust kernels or robust norms. A preliminary version of this work was presented as a conference paper. The paper discusses the problem of 3D point-set registration, which is a fundamental problem in computer and robot vision. The goal is to find the transformation that best aligns one of the point-sets to the other. The Iterative Closest Point (ICP) algorithm is the most well-known algorithm for efficiently registering two 2D or 3D point-sets under Euclidean (rigid) transformation. However, ICP is susceptible to local minima due to the non-convexity of the problem and the local iterative procedure