• 제어로봇시스템학회논문지
• /
• 제14권4호
• /
• pp.369-375
• /
• 2008

The kinematics and control problem of a visual alignment system is investigated, which plays a crucial role in the fabrication process of flat panel displays. The first solution is the inverse kinematics of a 4PPR parallel alignment mechanism. It determines the driving distance of each joint to compensate the misalignment between mask and panel. Second, an efficient vision algorithm for fast alignment mark recognition is suggested, where by extracting essential feature points to represent the geometry of a mark, the geometric template matching enables much faster object recognition comparing with the general template matching. Finally, the overall visual alignment process including the kinematic solution, vision algorithm, and joint control is implemented and experimental results are given.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2008년도 제38차 하계학술발표논문집 16권1호
• /
• pp.199-204
• /
• 2008

Visual surveillance is an active topic recently in Computer Vision. Event detection and recognition is one important and useful application of visual surveillance system. In this paper, we propose a new method to recognize the entering and exiting events based on the human's movement feature and the door's state. Without sensors, the proposed approach is based on novel and simple vision method as a combination of edge detection, motion history image and geometrical characteristic of the human shape. The proposed method includes several applications such as access control in visual surveillance and computer vision fields.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2013년도 추계학술대회
• /
• pp.8-11
• /
• 2013

We implement the foveation and frequency sensitivity feature of human visual system in discrete cosine transform (DCT) domain. Resolution of human visual perception decays as distance from the eye-focused point, known as foveation property, and the middle frequency components give most pleasant image quality to human than the low and high frequency components, which is the frequency sensitivity property of human visual system. For satisfying the foveation property, we enhanced the local contrast at the focused regions and smoothed local contrast at the non-focused regions in the DCT domain without bringing the blocking and ringing artifacts. Moreover, the energies at each DCT frequency components is modified with various degree to fulfill the frequency sensitivity property. The proposed method is verified by the subjective and objective evaluations that it can the improve the human perceptual visual quality.

• 한국컴퓨터정보학회논문지
• /
• 제24권5호
• /
• pp.11-17
• /
• 2019

In this study, we propose an ensemble engine, which is a framework for game engine optimized for visual novels genre, focusing on storytelling among various game genres. The game of Visual Nobel genre is based on multi-ending story and features branching of various scenarios according to user's choice. The proposed engine supports various multi-scenarios and multi-endings based on nodes according to the characteristics of these genres. In addition, it provides a convenient and intuitive user interface that not only enhances user immersion but also provides VR function to maximize the sense of presence. We will demonstrate the usefulness of the proposed game engine by designing the framework of a game engine suitable for this feature and actually creating variety stories automatically.

• 로봇학회논문지
• /
• 제16권3호
• /
• pp.238-244
• /
• 2021

Image mosaicking is a common and useful technique to visualize a global map by stitching a large number of local images obtained from visual surveys in underwater environments. In particular, visual inspection of underwater structures using underwater robots can be a potential application for image mosaicking. Feature-based pairwise image registration is a commonly employed process in most image mosaicking algorithms to estimate visual odometry information between compared images. However, visual features are not always uniformly distributed on the surface of underwater structures, and thus the performance of image registration can vary significantly, which results in unnecessary computations in image matching for poor-conditioned image pairs. This study proposes a pairwise registrability measure to select informative image pairs and to improve the overall computational efficiency of underwater image mosaicking algorithms. The validity and effectiveness of the image mosaicking algorithm considering the pairwise registrability are demonstrated using an experimental dataset obtained with a full-scale ship in a real sea environment.

• 한국항행학회논문지
• /
• 제19권2호
• /
• pp.133-139
• /
• 2015

이 논문에서는 영상정보를 기반으로 한 무인 시스템의 자율 항법기술에 대한 동향을 요약한다. 영상기반 항법기술로는 비주얼 서보잉, 비주얼 오도메트리, 영상 기반 SLAM(simultaneous localization and mapping)이 있다. 비주얼 서보잉은 목표 이미지와 현재 이미지 사이의 피쳐 차이로부터 원하는 속도 입력을 계산하여 무인 로봇을 목표 자세로 유도하는 데 사용된다. 비주얼 오도메트리는 무인 시스템이 영상정보를 바탕으로 자신의 이동 궤적을 추정하는 기술로, 기존의 dead-reckoning 방식보다 정확성을 향상시킬 수 있다. 영상 기반 SLAM은 무인 시스템이 영상 정보를 활용하여 미지의 환경에 대한 지도를 구축함과 동시에 자신의 위치를 결정해 나가는 기술로, 정확히 알지 못하는 환경에서 무인차량이나 무인기를 운용하는데 필수적이다. 이러한 기술들이 적용된 해외의 연구 사례들을 살펴봄으로써 영상기반 항법기술의 동향을 파악할 수 있었다.

• 정보처리학회 논문지
• /
• 제13권7호
• /
• pp.335-347
• /
• 2024

개방형 어휘 3차원 포인트 클라우드 개체 분할은 3차원 장면 포인트 클라우드를 훈련단계에서 등장하였던 기본 클래스의 개체들뿐만 아니라 새로운 신규 클래스의 개체들로도 분할해야 하는 어려운 시각적 작업이다. 본 논문에서는 중요한 모델 설계 이슈별 기존 모델들의 한계점들을 극복하기 위해, 새로운 개방형 어휘 3차원 개체 분할 모델인 Open3DME를 제안한다. 첫째, 제안 모델은 클래스-독립적인 3차원 마스크의 품질을 향상시키기 위해, 새로운 트랜스포머 기반 3차원 포인트 클라우드 개체 분할 모델인 T3DIS[6]를 마스크 제안 모듈로 채용한다. 둘째, 제안 모델은 각 포인트 세그먼트별로 텍스트와 의미적으로 정렬된 시각적 특징을 얻기 위해, 사전 학습된 OpenScene 인코더와 CLIP 인코더를 적용하여 포인트 클라우드와 멀티-뷰 RGB 영상들로부터 각각 3차원 및 2차원 특징들을 추출한다. 마지막으로, 제안 모델은 개방형 어휘 레이블 할당 과정동안 각 포인트 클라우드 세그먼트별로 추출한 2차원 시각적 특징과 3차원 시각적 특징을 상호 보완적으로 함께 이용하기 위해, 특징 앙상블 기법을 적용한다. 본 논문에서는 ScanNet-V2 벤치마크 데이터 집합을 이용한 다양한 정량적, 정성적 실험들을 통해, 제안 모델의 성능 우수성을 입증한다.

• The Journal of the Acoustical Society of Korea
• /
• 제28권3E호
• /
• pp.109-117
• /
• 2009

Identity recognition in real environment with a reliable mode is a key issue in human computer interaction (HCI). In this paper, we present a robust person identification system considering score-based optimal reliability measure of audio-visual modalities. We propose an extension of the modified reliability function by introducing optimizing parameters for both of audio and visual modalities. For degradation of visual signals, we have applied JPEG compression to test images. In addition, for creating mismatch in between enrollment and test session, acoustic Babble noises and artificial illumination have been added to test audio and visual signals, respectively. Local PCA has been used on both modalities to reduce the dimension of feature vector. We have applied a swarm intelligence algorithm, i.e., particle swarm optimization for optimizing the modified convection function's optimizing parameters. The overall person identification experiments are performed using VidTimit DB. Experimental results show that our proposed optimal reliability measures have effectively enhanced the identification accuracy of 7.73% and 8.18% at different illumination direction to visual signal and consequent Babble noises to audio signal, respectively, in comparison with the best classifier system in the fusion system and maintained the modality reliability statistics in terms of its performance; it thus verified the consistency of the proposed extension.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1996년도 한국자동제어학술회의논문집(국내학술편); 포항공과대학교, 포항; 24-26 Oct. 1996
• /
• pp.940-943
• /
• 1996

In teleoperating, as seeing the monitor screen obtained from a camera instituted in the working environment, human operator generally controls the slave arm. Because we can see only 2-D image in a monitor, human operator does not know the depth information and can not work with high accuracy. In this paper, we proposed a traded control method using an visual sensor for the purpose of solving this problem. We can control a teleoperation system with precision when we use the proposed algorithm. Not only a human operator command but also an autonomous visual sensor feedback command is given to a slave arm for the purpose of coincidence current image features and target image features. When the slave arm place in a distant place from the target position, human operator can know very well the difference between the desired image features and the current image features, but calculated visual sensor command have big errors. And when the slave arm is near the target position, the state of affairs is changed conversely. With this visual sensor feedback, human does not need coincide the detail difference between the desired image features and the current image features and proposed method can work with higher accuracy than other method without, sensor feedback. The effectiveness of the proposed control method is verified through series of experiments.

• 반도체디스플레이기술학회지
• /
• 제13권2호
• /
• pp.45-52
• /
• 2014

This paper proposes an efficient photo retrieval system that automatically indexes for searching of relevant images, using a combination of geo-coded information, direction/location of image capture device and content-based visual features. A photo image is labeled with its GPS (Global Positioning System) coordinates and direction of the camera view at the moment of capture, and the label leads to generate a geo-spatial index with three core elements of latitude, longitude and viewing direction. Then, content-based visual features are extracted and combined with the geo-spatial information, for indexing and retrieving the photo images. For user's querying process, the proposed method adopts two steps as a progressive approach, filtering the relevant subset prior to use a content-based ranking function. To evaluate the performance of the proposed scheme, we assess the simulation performance in terms of average precision and F-score, using a natural photo collection. Comparing the proposed approach to retrieve using only visual features, an improvement of 20.8% was observed. The experimental results show that the proposed method exhibited a significant enhancement of around 7.2% in retrieval effectiveness, compared to previous work. These results reveal that a combination of context and content analysis is markedly more efficient and meaningful that using only visual feature for image search.