Veri Analizi Araçları Kullanarak Eklemli Şekilleri Şekil Bileşenlerini Bulmadan Temsil Etme

2018-11-08
Genctav, Aslı
Tarı, Zehra Sibel
In this work, we propose a new representation for articulated shapes, which does not involve extracting shape components. We first form a high -dimensional feature space by collecting local shape information from each shape point and then perform data analysis to measure the distinctness of each shape point from the remaining points. The obtained distinctness values encode both local and global shape information. We represent shapes via histograms of the distinctness values computed via multiple shape scales.
Citation Formats
A. Genctav and Z. S. Tarı, “Veri Analizi Araçları Kullanarak Eklemli Şekilleri Şekil Bileşenlerini Bulmadan Temsil Etme,” 2018, p. 618, Accessed: 00, 2021. [Online]. Available: http://eng.harran.edu.tr/gap2018/wp-content/uploads/GAP2018_Bildiri_KitapciC49Fi28.12.18v2.pdf.