This paper describes algorithms for correlating images of arbitrary three-dimensional scenes by clusterizing correlated key points, using the Hough transform. The method is based on the well-known method of detecting objects, but an alternative approach is proposed for verifying clusters of correlated key points. Experimental results are given for different types of key points, confirming that the proposed method has a significant advantage over the use of the fundamental matrix.
© 2014 Optical Society of America
Original Manuscript: March 5, 2014
Published: June 19, 2014
R. O. Malashin, "Correlating images of three-dimensional scenes by clusterizing the correlated local attributes, using the Hough transform," J. Opt. Technol. 81, 327-333 (2014)