Objective: In order to solve the duality problem in pose estimation from single circle in the machine vision

an approach based on Euclidean angular constraints is presented to identify the unique pose result and the accuracy of pose estimation method is analyzed

which provides constructive suggestion on achieving accurate pose estimation from circles based on the experiments result. Method: First

the pose of single circle can be recovered from its projection with a calibrated camera

though the result is ambiguous; then

the unique pose result is indentified based on the Euclidean angular invariant which is a prior knowledge. Finally

the accuracy of the pose result is analyzed based on the theory of error propagation. Result: Experimental results indicate that the absolute angle error of circle plane is within 0.2 degree

the relative error of position determination is within 5‰

the absolute error of reconstructed distance between lines is within 8‰. Conclusion: It can accurately identify the unique pose from the ambiguous results from a single circle

the whole calculation process is concise and feasible to perform

the result is robust and reliable as well as a high accuracy.