• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.484-486
• /
• 2001

본 논문은 움직이는 거울을 사용하여 얻은 물체의 영상에서 물체의 3차원을 실시간에 복원하는 능동적 비전(active vision) 시스템을 제안한다. 거울을 사용한 기존의 스테레오 시스템에서는 거울들과 카메라가 고정되어 있어 물체의 부분적인 3차원 복원만이 가능한 반면, 제안하는 시스템은 거울에 의해 생성되는 두 개의 소실점을 이용함으로써 항상 실시간 카메라 보정이 가능하며, 간단하게 거울에 맺힌 영상들간의 대응점을 찾음으로써 물체의 3차원 복원이 가능하다.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.1
• /
• pp.13-20
• /
• 2016

In this paper, we propose an object recognition system that can effectively find out its category, its instance name, and several attributes from the color and depth images of an object with hierarchical feature learning. In the preprocessing stage, our system transforms the depth images of the object into the surface normal vectors, which can represent the shape information of the object more precisely. In the feature learning stage, it extracts a set of patch features and image features from a pair of the color image and the surface normal vector through two-layered learning. And then the system trains a set of independent classification models with a set of labeled feature vectors and the SVM learning algorithm. Through experiments with UW RGB-D Object Dataset, we verify the performance of the proposed object recognition system.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.10a
• /
• pp.285-288
• /
• 2007

A realization for automatic stock cubic measuring and distributing management embedded system with 3 dimensional ultrasonic sensing is presented in this paper. The height and width of cubic are measured by comparing the 3 values from 3 ultrasonic sensors with reference when an object is passing the conveyer and length is calculated by the passing time and velocity, compensate cubic values for error to vary with the environment temperature, and reduce the error by averaging the sensing data not to be right posture of object. The system enables to classify and load a packed stocks at the store and transportation practically based on the rectangular hexahedral objects.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.37 no.10
• /
• pp.1219-1227
• /
• 2013

A new type of wearable robot is developed for a disabled person with disabled limbs, that is, a person who cannot intentionally move his/her legs and arms. This robot can enable the disabled person to grip an object using eye movements. A gaze tracking algorithm is employed to detect pupil movements by which the person observes the object to be gripped. By using this gaze tracking 2D information, the object is identified and the distance to the object is measured using a Kinect device installed on the robot shoulder. By using several coordinate transformations and a matching scheme, the final 3D information about the object from the base frame can be clearly identified, and the final position data is transmitted to the DSP-controlled robot controller, which enables the target object to be gripped successfully.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.11 no.1
• /
• pp.1-15
• /
• 1986

Optic flow is 2D velocity projected on the image plane of 3D velocity of a moving surface element. In this paper, we survey techniques computing optic flows from an image time sequence of moving objects and techniques determining 3D velocities and surface structures of the moving objects from the optic flows determined.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.669-672
• /
• 2009

The techniques extract shape descriptors from 3D models and use these descriptors for indices for comparing shape similarities. Most similarity search techniques focus on comparisons of each individual 3D model from databases. However, our similarity search technique can compare not only each individual 3D model, but also partial shape similarities. The partial shape matching technique extends the user's query request by finding similar parts of 3D models and finding 3D models which contain similar parts. We have implemented an experimental partial shape-matching search system for 3D pagoda models, and preliminary experiments show that the system successfully retrieves similar 3D model parts efficiently.

• The KIPS Transactions:PartB
• /
• v.8B no.2
• /
• pp.201-207
• /
• 2001

3차원 물체표현은 컴퓨터 비젼 분야에서 중요한 역할을 차지하고 있다. 본 논문에서는 체적 표현법의 일종인 부품기반의 superquadric 모델을 통하여 3차원 기계 부품 물체를 표현하는 기법을 제안하였다. 이러한 부품기반의 superquadric 모델은 크기, 이동, 회전, 그리고 변형 등의 유한개의 계수들만을 가지고 다양한 3차원 체적소의 형상 표현이 가능하다는 장점이 있다. 따라서, superquadric 형상복구 과정을 통해서 이들 superquadric 계수들을 추출함으로써 3차원 단일 체적소 표현이 가능하다. 이때, 형상복구 과정의 입력은 3차원 거리 데이터로, 형상복구 과정은 3차원 nfcp를 이루는 각 체적소에 속하는 거리 데이터들을 입력으로하는 적합도 측정함수의 최소 자승법(LSM)에 의해 이루어진다. 이후에 3차원 물체 각 체적소에 해당하는 superquadric 계수들을 얻는다. 결과적으로 3차원 전체 물체에 해당하는 superquadric 모델은 이들 각 체적소에 해당하는 계수들의 집합으로 표현된다. 컴퓨터에서 합성한 합성영상과 실제 거리영상에 대한 실험을 통해 제안한 방법의 유용성을 입증하였다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.14 no.5
• /
• pp.243-248
• /
• 2014

This study presents an investigation into the ways to detect the areas of object movement with Kinect's Depth Frame, which is capable of receiving 3D information regardless of external light sources. Applied to remove noises along the boundaries of objects among the depth information received from sensors were the blurring technique for the x and y coordinates of pixels and the frequency filter for the z coordinate. In addition, a clustering filter was applied according to the changing amounts of adjacent pixels to extract the areas of moving objects. It was also designed to detect fast movements above the standard according to filter settings, being applicable to mobile robots. Detected movements can be applied to security systems when being delivered to distant places via a network and can also be expanded to large-scale data through concerned information.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.06a
• /
• pp.240-243
• /
• 2018

최근 VR(Virtual Reality)와 AR(Augmented Reality)의 발전에 따라 영상 또는 이미지에서 카메라와 물체 사이의 거리를 추정하는 기술에 대한 연구가 활발히 진행되고 있다. 본 논문에서는 카메라와 물체 사이의 거리 추정 방법 중에서 단일 카메라를 이용하여 촬영한 이미지의 흐림 정도를 분석하여 3D 거리를 추정하는 알고리즘을 연구한다. 특히 고가의 렌즈가 장착된 DSLR 카메라가 아닌 스마트폰 카메라 이미지에서 DFD를 이용한 거리 추정 방법 중 1개의 이미지를 이용한 3D 거리 추정 방법과 초점이 서로 다른 2개의 이미지를 결합하여 3D 거리를 추정하는 방법을 연구하고 최적회된 피사체 범위에 대해 연구하였다. 한 개의 이미지를 이용한 거리 추정에서는 카메라의 초점 거리를 200 mm로 설정할 때, 두 개의 이미지를 이용한 거리 추정에서는 두 이미지의 초점 거리를 각각 150 mm, 250 mm로 설정했을 때 가장 넓은 거리 추정 범위를 갖는다. 또한, 두 거리 추정 방법 모두 초점 거리가 가까울수록 가까운 물체의 거리 추정에 효율적인 것으로 나타났다.

• Journal of the Korea Computer Graphics Society
• /
• v.7 no.1
• /
• pp.27-35
• /
• 2001

Shadows are important elements in producing a realistic image. In rendering. generation of the exact shape and position of shadow is crucial in providing the user with visual cues on the scene. While the shadow map technique quickly generates a shadow for the scene wherein objects and light sources are fixed. it gets slow down as they start to move. In this paper. we apply an image-based rendering technique to generate shadows in real-time using graphics hardware. Due to the heavy requirement of storage for a shadow map repository. we use a wavelet-based compression scheme for effective compression. Our method will be efficiently used in generating realistic scenes in many real-time applications such as 3D games and virtual reality systems.