Abstract
깊이 영상을 이용하는 추정기 학습 방법 및 자세 추정 방법이 개시된다. 일 실시예에 따른 추정기 학습 방법은 합성 데이터와 실제 데이터 사이의 연관 관계에 기초하여, 객체의 자세를 추정하기 위한 추정기를 학습시킬 수 있다. 또한, 일 실시예에 따른 자세 추정 방법은 학습된 추정기를 이용하여 객체의 자세를 추정할 수 있다.
Machine Translation: Disclosed are the estimator learning method for using the depth image and the attitude estimation method. The estimator learning method according to the embodiment can cultivate the estimator for estimating the posture of the object based on the associative relationship between actual data and synthetic data. Moreover, the posture of the object can be estimated using the estimator in which the attitude estimation method according to the embodiment is learned
Machine Translation: Disclosed are the estimator learning method for using the depth image and the attitude estimation method. The estimator learning method according to the embodiment can cultivate the estimator for estimating the posture of the object based on the associative relationship between actual data and synthetic data. Moreover, the posture of the object can be estimated using the estimator in which the attitude estimation method according to the embodiment is learned
Translated title of the contribution | Estimator Learning Method and Pose Estimation Method Using a Depth Image |
---|---|
Original language | Korean |
Patent number | KR101758064000 |
IPC | G06T 7/20 and G06T 7/00 |
Priority date | 5/06/13 |
Publication status | Published - 14 Jul 2017 |