A New Corner Detector Approach for Occupancy Grid Map Merging A New Corner Detector Approach for Occupancy Grid Map Merging

Resumen

The problem of merging maps in SLAM is one of the most studied in this field, because it allows to extend the SLAM algorithms to Multi-SLAM field. This issue is treated as a problem of merging images. In computer vision, merging images issue, nowadays it has many approaches for solving it, using feature extractors and descriptors. In this paper, we propose and show a new corner detector technique that can be used in map images. The results obtained in our tests show that, our corner detector is reliable and efficient to extract features in images generated by SLAM algorithms. Furthermore, we compared our algorithm with others feature detectors like Harris Corner Detector, Shi-Tomasi Detector, among others. We found out our corner detector has a good and reliable performance, doing the extraction from those kinds of map images.