2D Rotation-Angle Measurement Utilizing Least Iterative Region Segmentation.

When geometric moments are used to measure the rotation-angle of plane workpieces, the same rotation angle would be obtained with dissimilar poses. Such a case would be shown as an error in an automatic sorting system. Here, we present an improved rotation-angle measurement method based on geometric moments, which is suitable for automatic sorting systems. The method can overcome this limitation to obtain accurate results. The accuracy, speed, and generality of this method are analyzed in detail. In addition, a rotation-angle measurement error model is established to study the effect of camera pose on the rotation-angle measurement accuracy. We find that a rotation-angle measurement error will occur with a non-ideal camera pose. Thus, a correction method is proposed to increase accuracy and reduce the measurement error caused by camera pose. Finally, an automatic sorting system is developed, and experiments are conducted to verify the effectiveness of our methods. The experimental results show that the rotation angles are accurately obtained and workpieces could be correctly placed by this system.