An automatic grid corner extraction technique for camera calibration

Lixia Yang, Chao Tian, Vinay A. Vaishampayan, Amy R. Reibman

Abstract: Camera calibration is essential for many computer vision and image processing applications. However, this calibration process can be rather time consuming and may require a significant amount of human intervention. Calibration models traditionally employ a calibration grid whose four corner points must be marked by hand on a per-frame basis. The objective of this work is to develop a technique for processing these frames rapidly, with as little human intervention as possible. We propose an algorithm to extract the boundaries of the calibration grid automatically, based on a spectral analysis of HD (high-definition) video frames. The accuracy of the intrinsic parameters estimated using our automatic method is evaluated through comparison with those obtained using a method that requires hand labeling of the corner points.