Naslov
On the Segmentation of Scanned 3D Human Body Models

Autori
Ðonlić, Matea ; Petković, Tomislav ; Peharec, Stanislav ; Berryman, Fiona ; Pribanić, Tomislav

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
Abstract Book of Seventh Croatian Computer Vision Workshop (CCVW 2018) / Lončarić, Sven ; Petković, Tomislav - Zagreb : Fakultet elektrotehnike i računarstva Sveučilišta u Zagrebu, 2018, 17-17

Skup
Seventh Croatian Computer Vision Workshop (CCVW 2018)

Mjesto i datum
Zagreb, Hrvatska, 24.09.2018

Vrsta sudjelovanja
Poster

Vrsta recenzije
Nije recenziran

Ključne riječi
3D scanner ; human body models ; segmentation

Sažetak
Analysing the human body in three-dimensional (3D) space is becoming more and more popular due to the large number of available 3D scanning solutions. Using 3D body models enables detection of many different body parameters which cannot be assessed using 2D images only. Once a 3D body model is generated, it is usually segmented so that specific body parts can be further analysed. In this paper we propose a new approach to the segmentation of the human body. Our anthropometry-based approach can successfully segment various body-types into six main segments – head, torso, left and right arm, and left and right leg. The proposed method has been tested successfully on standard benchmark models and on 3D point cloud data acquired from a low- cost 3D scanner.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Računarstvo

POVEZANOST RADA