• 한국전자통신학회논문지
• /
• 제19권3호
• /
• pp.633-640
• /
• 2024

이 연구는 헬스케어 분야에서 활용 가능한 AI 기반의 3D 체형 모델링 기술을 개발하고, 이를 통해 사용자의 체형 변화와 건강 상태를 모니터링 할 수 있는 시스템을 제안한다. 사이즈코리아의 데이터를 활용하여 2D 이미지로부터 3D 체형 이미지를 생성하는 모델을 개발하고, 다양한 모델을 비교하여 가장 성능이 우수한 모델을 선정하였다. 최종적으로, 개발된 기술을 통해 개인 맞춤형 건강 관리, 운동 추천, 식단 제안 등의 시스템 프로세스를 제안함으로써 질병 예방 및 건강 증진에 기여하고자 하였다.

• 한국의류산업학회지
• /
• 제22권1호
• /
• pp.87-93
• /
• 2020

This study provide basic data to develop a dress form reflecting body shape characteristics by age and to produce a 3D body form in a virtual fitting program. A comparative analysis was conducted on the size, section shape, and slope of side shape of the modeling form by the sizing of the basic female avatar in CLO 3D, one of the 3D apparel CAD programs and the body form of women in their 20s-50s by body shape in the 6th Size Korea (2010). First, all the differences were formed similarly in the direct measurements between the 3D avatar and the body form were within 1 mm. Second, in a comparison of the section form of the avatar and body form in Size Korea, the avatar was formed in straight body shape and did not reflect a spinal curvature according to age. As a result of an examination of the items with a difference over 5° in the slope of side shape, there were angle differences in numerous body shapes in the angles of the side upper abdomen, side upper back, and side upper bust, and the avatar's bust shape was expressed more flatly compared to body form. It will be possible to produce an avatar that can adequately reflect body shape characteristics by adding detailed length and angle items by the region like waist back length and front length in producing the avatar reflecting body shape characteristics, instead of a standard body shape.

• 패션비즈니스
• /
• 제19권3호
• /
• pp.130-143
• /
• 2015

By studying 3-D virtual human modeling, motion-capturing and clothing simulation for easier and safer work clothes development, this research aimed (1) to categorize heavy manufacturing work motions; (2) to generate a 3-D virtual male model and establish painting work motions within a 3-D virtual clothing simulation system through computerized body scanning and motion-capturing; and finally (3) to suggest simulated clothing images of painting work clothes developed based on virtual male avatar body measurements by implementing the work motions defined in the 3-D virtual clothing simulation system. For this, a male subject's body was 3-D scanned and also directly measured. The procedures to edit a 3-D virtual model required the total body shape to be 3-D scanned into a digital format, which was revised using 3-D Studio MAX and Maya rendering tools. In addition, heavy industry workers' work motions were observed and recorded by video camera at manufacturing sites and analyzed to categorize the painting work motions. This analysis resulted in 4 categories of motions: standing, bending, kneeling and walking. Besides, each work motion category was divided into more detailed motions according to sub-work posture factors: arm angle, arm direction, elbow bending angle, waist bending angle, waist bending direction and knee bending angle. Finally, the implementation of the painting work motions within the 3-D clothing simulation system presented the virtual painting work clothes images simulated in a dynamic mode.

• 한국의상디자인학회지
• /
• 제25권4호
• /
• pp.147-161
• /
• 2023

This study used 3D anthropometric data from the 8th Size Korea to type and analyze whole body shapes of obese women in their 20s and 30s, and constructed dimensional data for human body items needed to create a 3D human body model for each type. The data analysis used data from 148 obese women in their 20s and 30s, and a total of 48 index values, drop values, and angle items were subjected to factor analysis and one-way variance analysis to categorize body types and verify significant differences by type. As a result of the factor analysis, 12 factors were extracted and divided into 4 body types. Type 1 is a 'standard type with a curved torso with balanced upper and lower body lengths', Type 2 is a 'bending forward type with a short, thick lower body, and an uncurved torso', Type 3 is a 'lean back type with a long and thin lower body and an H-shape torso', Type 4 is a 'sway back type with a long and thick lower body and abdominal obesity'. The representative body type of obese women in their 20s and 30s was identified as Type 1. The constructed body shape information will be used as basic data for future 3D human body modeling.

• 한국컴퓨터그래픽스학회논문지
• /
• 제28권4호
• /
• pp.1-11
• /
• 2022

선형 스키닝 기반 3D 인체 생성 모델 SMPL (Skinned Multi-Person Linear Model)은 1990년대 미국 거주 미국인들 3천 8백여명을 대상으로 만들어진 3D 신체 데이터베이스인 CEASAR 로부터 최적화 기법을 통해 만들어진 가장 널리 쓰이는 통계적인 3D 모델이다. 본 연구는 한국인 체형의 통계적 특성을 보다 정확히 표현하는 SMPL기반의 한국인 체형 3D 모델을 제안한다. 이를 위해 우리는 한국인 여성 2천7백여명의 신체 각 부위의 실측 데이터에 기존 3D SMPL 모델을 피팅하는 비선형 최적화 알고리즘을 개발한다. 이를 사용하여 한국인 3D 신체 데이터베이스를 구축하고, 주성분 분석 방법으로 한국인 체형 기반 매개화된 3D 모델을 개발한다. 본 연구를 통해 제안하는 한국인의 체형적 특징을 가진 블렌드쉐입과 새로운 체형 파라미터는 기존 모델이 표현하는 체형에 비해 한국인 체형 데이터 특성을 잘 반영함을 확인하였다. 뿐만 아니라, 우리의 모델은 SMPL에 비해 신체 실측 데이터에 대한 피팅 정확도를 개선함을 확인하였다. 제안된 모델은 향후 아바타 생성이나 인체 형상 측정 등 다양한 용도로 사용될 수 있다.

• 한국CDE학회논문집
• /
• 제14권6호
• /
• pp.364-373
• /
• 2009

In this paper, we suggest the method for constructing parameterized human body model which has any required body sizes from 3D scan data. Because of well developed 3D scan technology, we can get more detailed human body model data which allow to generate precise human model. In this field, there are a lot of research is performed with 3D scan data. But previous researches have some limitations to make human body model. They need too much time to perform hole-filling process or calculate parameterization of model. Even more they missed out verification process. To solve these problems, we used several methods. We first choose proper 125 3D scan data from 5th Korean body size survey of Size Korea according to age, height and weight. We also did post process, feature point setting, RBF interpolation and align, to parameterize human model. Then principal component analysis is adapted to the result of post processed data to obtain dominant shape parameters. These steps allow to reduce process time without loss of accuracy. Finally, we compare these results and statistical data of Size Korea to verify our parameterized human model.

• Journal of Computational Design and Engineering
• /
• 제3권2호
• /
• pp.121-131
• /
• 2016

Organic modeling of 3D characters is a challenging task when it comes to correctly modeling the anatomy of the human body. Most sketch based modeling tools available today for modeling organic models (humans, animals, creatures etc) are focused towards modeling base mesh models only and provide little or no support to add details to the base mesh. We propose a hybrid approach which combines geometrical primitives such as generalized cylinders and cube with Shape-from-Shading (SFS) algorithms to create plausible human character models from sketches. The results show that an artist can quickly create detailed character models from sketches by using this hybrid approach.

• 대한가정학회지
• /
• 제44권2호
• /
• pp.127-135
• /
• 2006

This paper suggests guidelines for virtual clothes modeling and draping software suitable for clothes designers. We first analyze Maya Cloth, which is widely used in game and animation fields, and which has been adopted by Pad System as a 3D cloth draping system. We then discuss what functions and procedures would improve Maya Cloth to assist designers in being able to create the clothes they have conceptualized. While Maya Cloth has many good functions and features forvirtual cloth modeling and draping, it treats input 2D patterns as approximat and it creates 3D clothes by considering other factors such as the 3D body model. As a result, it is hard for clothes designers to control the shape of the 3D clothes by changing 2D patterns. Furthermore, Maya Cloth does not handle seamlines satisfactorily. We suggest that the following new features should be added to Maya Cloth : respecting the input 2D patterns, handling seamlines, and controlling the shape of the clothes in 3D space.

• 방송공학회논문지
• /
• 제27권5호
• /
• pp.638-653
• /
• 2022

본 논문은 고화질 2차원 전신 영상을 입력으로 받아 영상 속 인물이 입고 있는 의상 패턴과 체형 정보를 추정한 후, 이를 반영한 3차원 가상 휴먼의 생성 기법을 제안한다. 의상의 패턴을 얻기 위해서 Cascade Mask R-CNN을 이용하여 의상 분할을 진행한다. 이후 Pix2Pix로 경계를 블러 및 배경색을 추정하고, UV-Map 기반으로 변환하여 3차원 의상 메쉬의 UV-Map을 얻을 수 있다. 또한, SMPL-X를 이용하여 체형 정보를 얻고 이를 기반으로 의상과 신체의 기본 메쉬를 변형한다. 앞서 얻은 의상 UV-Map, 체형이 반영된 의상과 신체의 메쉬를 이용해 최근 각광받고 있는 게임 엔진인 언리얼 엔진에서 렌더링하여 최종적으로 사용자가 그의 외형이 반영된 3차원 가상 휴먼의 애니메이션을 볼 수 있도록 한다.

• 한국의류산업학회지
• /
• 제24권4호
• /
• pp.451-459
• /
• 2022

This study aimed to develop a a custom-made dress form for draping using a live model's 3D body scan obtained from an entry-level 3D handheld scanners, 3D modeling software and 3D printing technology. A female subject was recruited whose body size fell under the normal (N) body shape criteria suggested by KS K 0051. First, the handheld scanner reduced the length of the legs in scanning, but most of the scanning operations between the neck and crotch levels were conducted accurately. Therefore, this study was designed to develop a torso dress form. The full body 3D scan was edited into a torso shape using ZBrush® software. Using Rhinoceros^® and Materialise's Magics software, a 3D body scan was modeled so that the user could fit two types of mannequin stands (one with a neck fixation from above and one with an insert from below) to the dress form. The body scan was divided into 9 pieces to fit the printable size of the Stratasys 3D printer Fortus 250mc, and the cross-sectional distance from the center to the periphery was downsized by 2 mm. After outputting the dress form scan file with a 3D printer, the dress form was manufactured by the first covering it with a 4 oz nonwoven pad and the second covering with a single jersey material.