Oliver Sidla, Michal Kottman, Vanda Benesova
Abstract. In this work we have tested several keypoint/feature descriptor combinations for logo detection in order to estimate theÂ robustness of different algorithms with respect to logo type and image quality. Our tests have shown that the wellÂ established SURF/SURF combination seems to perform best, followed by Calonderâ€™s keypoint detector/Random FernÂ combination. The authors still believe that the LDETECTOR/Fern combination, especially when using the compressedÂ signature framework is used (which has not yet been implemented for this test) bears a large potential. Its speedÂ advantage and ability for fast online learning should make it an interesting alternative to SURF/SURF.