An accelerative orthogonal iteration algorithm about the classical and wildly used orthogonal iteration algorithm for camera pose estimation is proposed for real time computation. The key idea is to integrate the steps in each iteration. The repetitive computation in each iteration can be abstracted and done before iteration. The computational complexity of each iteration is reduced from O(n) to O(1). So that, more iteration can be done in short time, and the accuracy is improved as well. The contrastive simulation and real data experiments show the efficiency and accuracy of the accelerative algorithm. Experimentally, the accelerative algorithm with the robust perspective n point (RPnP) initialization has nearly the same accuracy as maximum likelihood estimation (MLE), and is the fastest algorithm when there are few control points.