• Journal of Multimedia Information System
• /
• 제9권3호
• /
• pp.183-190
• /
• 2022

Image-based 3D object detection is one of the important and difficult problems in autonomous driving and robotics, and aims to find and represent the location, dimension and orientation of the object of interest. It generates three dimensional (3D) bounding boxes with only 2D images obtained from cameras, so there is no need for devices that provide accurate depth information such as LiDAR or Radar. Image-based methods can be divided into three main categories: monocular, stereo, and multi-view 3D object detection. In this paper, we investigate the recent state-of-the-art models of the above three categories. In the multi-view 3D object detection, which appeared together with the release of the new benchmark datasets, NuScenes and Waymo, we discuss the differences from the existing monocular and stereo methods. Also, we analyze their performance and discuss the advantages and disadvantages of them. Finally, we conclude the remaining challenges and a future direction in this field.

• International Journal of Air-Conditioning and Refrigeration
• /
• 제8권2호
• /
• pp.11-22
• /
• 2000

Holographic interferometric tomography can provide reconstruction of instantaneous three-dimensional gross flow fields. The technique however confronts ill-posed reconstruction problems in practical applications. Experimental data are usually limited in projection and angular scanning when a field is captured instantaneously or under the obstruction of test models and test section enclosures. An algorithm, based on a series expansion method, has been developed to improve the reconstruction under the ill-posed conditions. A three-dimensional natural convection flow around two interacting isothermal cubes is experimentally investigated. The flow can provide a challenging reconstruction problem and lend itself to accurate numerical solution for comparison. The refractive index fields at two horizontal sections of the thermal plume with and without an opaque object are reconstructed at a limited view angle of 80$\circ$. The experimental reconstructions are then compared with those from numerical calculation and thermocouple thermometry. It confirms that the technique is applicable to reconstruction of reasonably complex, three-dimensional flow fields.

• 대한인간공학회:학술대회논문집
• /
• 대한인간공학회 1996년도 춘계학술대회논문집
• /
• pp.168-185
• /
• 1996

In this research, an investigation of the eye movement was performed when the objects perceived by the eye are three-dimensional objects, two dimensional images, and three dimensional image. This investigation was done by observing the EOG(Electro-oculogram) waves which were achived from experiments. The observed waves were categorized into several groups. Differences among the waves were analyzed for each object and image perceived by the eye. In order to obtain waves to be analyzed, two kinds of experiments were performed. In each experiments, the saccadic eye movement and the smooth pusuit eye movement was considered as an independent variable individually. Waves obtained from the experiments were categorized into 4 types(Wave1, Wave2, Wave3, Wave4) depending on their characteristics. Unlike from the other images, three dimensional image was revealed as a key factor for the active movement of the eye. Futhermore, a unique eye movement was observed in the case of three dimensional image where the focus of the eye was achived in three steps.

• 한국한의학연구원논문집
• /
• 제1권1호
• /
• pp.273-287
• /
• 1995

Moire topography, a simple technique for three-dimensional quantitation, was used to provide interference fringe photographs of the human back with sufficient accuracy to be used for detecting patient with asymmetry due to scoliosis, the disease of cervix and lumbar, muscle dysfunction. Contour lines are a suitable and widely accepted method of describing a three-dimensional surface. In the moire technique, contour lines of an object are produced as interference fringes while the object is illuminated by a spotlight through a special grating. The fringe pattern is produced by the interference of the grating and its shadow on the object. A photograph of a moire pattern on the human back will permit an assessment of the overall body shape and the symmetry of the back. This study uses shadow moire topography. Moire topography provides a non-invasive technique for quantifying the shape of the human body. In the use of moire topography for the Oriental Medicine Diagnosis, the strength of moire lies in the ablility to detect change due to deformity of human body.

• 한국광학회지
• /
• 제17권6호
• /
• pp.525-531
• /
• 2006

3차원 물체의 합성된 컴퓨터 형성 홀로그램(CGH)을 새로운 형태의 3차원 좌표변환 기술을 사용하여 제작하였다. 제작될 3차원 물체는 CCD 카메라를 사용하여 시야각도(viewing angle)를 변화시키면서 여러개의 영상으로 기록하였다. 이렇게 얻어진 영상 데이터는 2차원 복소 매트릭스를 사용하여 수학적인 방법으로 계산되고, 최종적으로 합성된 3차원 CGH로 부호화하였다.

• Journal of the Optical Society of Korea
• /
• 제5권3호
• /
• pp.99-109
• /
• 2001

This paper presents a phase-only modulation scheme for a three-dimensional (3-D) image matching system to improve optical efficiency of the system. The 3-D image matching system is based on the two mask heterodyne scanning. A hologram of the 3-D reference object is first created and then the phase of the hologram is extracted. The phase of the hologram is represented as one mask with the other mask being a plane wave. The superposition of each beam modulated by the two masks generated a scanning beam pattern. This beam pattern scans the 3-D target object to be recognized. The output of the scanning system gives out the correlation of the phase-only hologram of the reference object and the complex hologram of the target object. Since a hologram contains 3-D information of an object as a form of fringe pattern, the correlation of holograms matches whole 3-D aspect of the objects. Computer simulations are performed with additive gaussian noise and without noise for the complex hologram modulation scheme and the phase-only hologram modulation scheme. The computer simulation results show that the phase-only hologram modulation scheme improves the optical efficiency. Thus the system with the phase-only hologram modulation scheme is more robust than the system with the complex hologram modulation scheme.

• 대한인간공학회지
• /
• 제16권1호
• /
• pp.1-13
• /
• 1997

In this research, and investigation of the eye movement was performend when the objects perceived by the eye are three-dimensional objects, two dimensional images, and trhee dimensional images. This investigation was done by observing the EQG(Electro- oculogram) waves which were obtained from experiments. The observed waves were categorized into several groups. Differences among the waves were analyzed for each object and image perceived by the eye. In order to obtain waves for analysis, two kinds of experiments were performen. In each experiment, the saccadic eye movement and the smooth pursuit eye movement were considered as an independent variables. Waves obtained from the experiments were categorized into 4 types (Wave-1, Wave-2, Wave-3, Wave-4) depending on their characteristics. Unlike the other images, three dimensional image was revealed as a key factor for the active movement of the eye. Futhermore, a unique eye movement was observed in the case of three dimensional images where the focus of the eye was achieved through three steps.

• Spatial Information Research
• /
• 제15권1호
• /
• pp.35-52
• /
• 2007

실세계에 가까운 가상현실의 표현기술과 웹을 통해 공간정보를 제공하는 3차원 GIS는 새로이 주목받는 분야 중에 하나이다. 특히, 공간데이터의 상호운용성에 대한 관심이 높아지고 있는 가운데, OGC는 상호운용을 지원하는 공간객체의 위상관계 명세를 발표하였다. 그러나 이 명세서는 2차원 공간객체에 국한되어있다. 이에 본 연구에서는 도시의 경관개선과 GIS 활용 기반 조성의 측면에서 3차원 공간객체의 위상관계를 구축하였다. 나아가 이를 토대로 신도시의 실정에 맞는 경관정보 가시화 방안을 제시하였다. 결과적으로 장소에 구애받지 않고 상시접속이 가능한 현실감 있는 도시경관에 대한 정보공유의 기틀을 마련하였다는 점에서 보다 큰 의의가 있다고 할 것이다.

• 한국음향학회지
• /
• 제19권3호
• /
• pp.20-25
• /
• 2000

본 논문에서는 3차원 물체 인식을 위한 표면 분류 및 임계치 선정 방법에 대해 제안 하고자 한다. 3차원 영상 처리는 크게 거리 영상의 획득과 특징 추출 그리고 정합 과정으로 이루어진다. 본 논문에서는 전체 3차원 영상 처리 시스템중 거리 영상을 입력으로 했을 시 형상 특징을 추출하는 방법에 대해 제안하고자 한다. 이를 위해 첫째, 거리 영상의 깊이 변화 부호 값의 분포 특성에 따라 표면을 분류하는 방법을 제안하고자 한다. 또한 평균 곡률과 가우스 곡률을 이용하여 표면을 분류했던 기존 방법을 토대로 그의 문제점이었던 실제 거리 영상에서의 임계치 선정 방법에 대하여 제안하고자 한다. 끝으로 제안한 방법의 유용성을 실험에 의해 입증하고자 한다.

• 대한전자공학회논문지
• /
• 제27권1호
• /
• pp.151-163
• /
• 1990

In this paper, rule-based stereo algorithm is explored to obtain three dimensional information of an object. In the preprocessing of the stereo matching, feature points of stereo images must be less sensitive to noise and well linked. For this purpose, a new feature points detection algorithm is developed. For performing the stereo matching which is most important process of the stereo algorithm, the feature representation of feature points is first described. The feature representation is then used for a rule-based stereo algorithm to determine the correspondence between the input stereo images. Finally, the three dimensional information of the object is determined from the correspondence of the feature points of right and left images.