作者: Virginio Cantoni , Marco Ferretti
DOI: 10.1007/978-1-4615-2413-7_2
关键词:
摘要: In order to achieve the high performance that real applications require, correct computation on relevant image data at right time is essential. Following studies vision and perception in humans, two phases can be distin- guished: (1) a preattentive phase, which visual system only dedicated detection of events regions interest within its wide field view, (2) an attentive extensive analysis restricted amount performed. Correspondingly, equivalent computational paradigm will introduced reduce huge raw transduced by standard artificial sensor. Such provides for use variable-resolution grids, according detail required task, thus obtaining multiresolution systems with different-sized layers.