“…Max-trees allow the efficient implementation of many connected filters like area-open (Vincent, 1994), hmax (Salembier et al, 1998), ultimate opening (Fabrizio and Marcotegui, 2009), statistical attribute filters (Teeninga et al, 2015), vector attribute filters (Kiwanuka and Wilkinson, 2015), among others. They have been used in a wide range of automatic applications like object recognition (Souza et al, 2014), scale and rotation invariant image classification (Urbach et al, 2007), detection, tracking and recognition of license plates (Donoser et al, 2007), recognition of text in natural scenes (Merino-Gracia et al, 2012), 3D magnetic resonance (MR) brain segmentation (Dokládal et al, 2003), angiographic image segmentation (Caldairou et al, 2009), dermatological image segmentation (Naegel et al, 2007), medical image registration (Richard et al, 2000), remote sensing (Ghamisi et al, 2016b;, detection of local features like Maximally Stable Extremal Regions (MSER) (Matas et al, 2002) and Morse Regions (Xu et al, 2014), among many others.…”