Bilateral Symmetry Detection in Perspective Based on Vanish Features

Symmetry detection is an important research field in image analysis and computer vision. It plays an important role in object match, recognition and location. However, the change of perspective projection angle increases the difficulty of symmetry detection. Therefore, it is of great significance to detect the symmetry of objects under perspective projection. Current feature point-based methods mostly use multiple sets of matched point pairs to detect the symmetry axis line, which not only introduces massive calculation but also interference between feature point pairs. Hence, we propose a novel feature point match-based symmetry axes lines detection method of planar graphs in perspective. This approach detects matched pairs based on the BEBLID descriptor and AdaLAM mismatch removal, and adopts binocular cameras to get a potential symmetry axis line from a pair of matched points, avoiding the influence of other matched points. Experimental results show that this method can achieve accurate and efficient detection of symmetry axis.