• 방송공학회논문지
• /
• 제28권2호
• /
• pp.230-237
• /
• 2023

자율주행은 4차 산업의 핵심 기술로 차, 드론, 자동차, 로봇 등 다양한 곳에 응용 가능하다. 그 중 위치 추정 기술은 GPS, 센서, 지도 등을 활용하여, 객체나 사용자의 위치를 파악하는 기술로 자율주행을 구현하기 위한 핵심적인 기술 중 하나이다. GPS나 LIDAR 등의 센서를 이용하여 위치 추정이 가능하지만, 이는 매우 고가이고 무거운 장비를 탑재해야 하며 지하 혹은 터널 등 전파 방해가 있는 곳의 경우 정밀한 위치 추정이 어렵다는 단점이 있다. 본 논문에서는 이를 보완하기 위해 저가의 비전 카메라로 획득한 컬러 영상을 입력으로 하여 관심 영역 추출 네트워크와 영상 분할 지도를 이용한 영상 검색 기술을 제안한다.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2004년도 춘계학술발표회논문집
• /
• pp.275-278
• /
• 2004

The geo-referencing of satellite imagery is a key task in remote sensing. GCPs are points the position of which is known both in the image and in the supporting maps. Mapping function makes the determination of map coordinates of all image pixels possible. Generally manual operations are done to identify image points corresponding to the points on a digital topographic map. In order to accurately measure ground coordinates of GCPs, differential global positioning system (DGPS) surveying are used. To acquire the sufficient number of well distributed GCPs is one of the most time-consuming and cost-consuming tasks. This paper describes the procedure of automatically extracting GCOs using GCP database. GCP image chips and image matching technique are used for automatic extraction of GCPs. We developed image processing tool for making image chip GCPs and Web Server for management of GCPs.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2000년도 추계학술대회 논문집
• /
• pp.934-938
• /
• 2000

This paper presents a laser cutting of 2D image. 2D image in pixel graphic format is converted into vector graphic image by image processing. Bitmap graphics are made easily, but can not being used in application works for geometry transition. The Sobel's Edge detection method is used to find boundary points on 2D image. The points are fitted into curves with sampling and filtering. Sampling can provide efficient computation and filtering reconstuct features in image. The NC code is generated on MURBS curve of the points. Also, the offset of contour and cutting conditions are considered.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권12호
• /
• pp.3149-3165
• /
• 2013

Content based image retrieval is increasingly gaining popularity among image repository systems as images are a big source of digital communication and information sharing. Identification of image content is done through feature extraction which is the key operation for a successful content based image retrieval system. In this paper content based image retrieval system has been developed by adopting a strategy of combining multiple features of shape, color and relevance feedback. Shape is served as a primary operation to identify images whereas color and relevance feedback have been used as supporting features to make the system more efficient and accurate. Shape features are estimated through second derivative, least square polynomial and shapes coding methods. Color is estimated through max-min mean of neighborhood intensities. A new technique has been introduced for relevance feedback without bothering the user.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권10호
• /
• pp.4968-4986
• /
• 2017

In this paper, we primarily address the difficulty of automatic generation of a plausible depth map from a single image in an unstructured environment. The aim is to extrapolate a depth map with a more correct, rich, and distinct depth order, which is both quantitatively accurate as well as visually pleasing. Our technique, which is fundamentally based on a preexisting DepthTransfer algorithm, transfers depth information at the level of superpixels. This occurs within a framework that replaces a pixel basis with one of instance-based learning. A vital superpixels feature enhancing matching precision is posterior incorporation of predictive semantic labels into the depth extraction procedure. Finally, a modified Cross Bilateral Filter is leveraged to augment the final depth field. For training and evaluation, experiments were conducted using the Make3D Range Image Dataset and vividly demonstrate that this depth estimation method outperforms state-of-the-art methods for the correlation coefficient metric, mean log10 error and root mean squared error, and achieves comparable performance for the average relative error metric in both efficacy and computational efficiency. This approach can be utilized to automatically convert 2D images into stereo for 3D visualization, producing anaglyph images that are visually superior in realism and simultaneously more immersive.

• Current Optics and Photonics
• /
• 제4권4호
• /
• pp.286-292
• /
• 2020

Imaging through multicore fiber (MCF) is of great significance in the biomedical domain. Although several techniques have been developed to image an object from a signal passing through MCF, these methods are strongly dependent on the surroundings, such as vibration and the temperature fluctuation of the fiber's environment. In this paper, we apply a new, strong technique called deep learning to reconstruct the phase image through a MCF in which each core is multimode. To evaluate the network, we employ the binary cross-entropy as the loss function of a convolutional neural network (CNN) with improved U-net structure. The high-quality reconstruction of input objects upon spatial light modulation (SLM) can be realized from the speckle patterns of intensity that contain the information about the objects. Moreover, we study the effect of MCF length on image recovery. It is shown that the shorter the fiber, the better the imaging quality. Based on our findings, MCF may have applications in fields such as endoscopic imaging and optical communication.

• Journal of Information Display
• /
• 제13권2호
• /
• pp.83-89
• /
• 2012

The crosstalk evaluation of multiview autostereoscopic three-dimensional (3D) displays is discussed, with both the human and technical factors investigated via image quality assessment. In the imaging performance measurements and analysis for a multiview autostereoscopic display prototype equipment, it was inferred that crosstalk would have both a positive and a negative effect on the imaging performance of the equipment. The importance of the attached diaphragm in the crosstalk evaluation was proposed and then experimentally verified, using the developed prototype equipment. The luminance distribution and crosstalk situation were given, with two different diaphragm arrays applied. The analysis results showed that the imaging performance of this 3D display system can be improved with minimum changes to the system structure.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2006년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.144-146
• /
• 2006

The Data Matrix of two-dimensional bar codes is a new technology capable of holding relatively large amounts of data compared to the conventional one-dimensional bar code which is just a key that can access detailed information to the host computer database. A secret key is used to prevent a watermark from malicious attacks. We encoded copyright information into a Data Matrix bar code for encoding process and it was spread a pseudo random pattern using owner key. We embedded a randomized watermark into the image using watermark's embedding position, pattern generated with a secret key. The experimental results have shown that the proposed scheme has good quality and is very robust to various attacks, such as JPEG compression and noise. Also the performance of the proposed scheme is verified by comparing the copyright information with the information which is extracted from a bar code scantier.

• International Journal of Costume and Fashion
• /
• 제2권
• /
• pp.73-85
• /
• 2002

• 한국컴퓨터정보학회논문지
• /
• 제16권6호
• /
• pp.155-164
• /
• 2011

스마트폰을 기반으로 한 앱은 병원의 의료 서비스, 은행 및 카드사의 금융서비스, 기업 및 가정의 유비쿼터스 기술 등 다양한 영역에 활용되고 있다. 이러한 서비스 환경에서 외부인에 의한 스마트폰의 불법적인 노출은 공 사적 정보의 유출을 포함한 자산의 손실이 발생한다. 이를 위한 보호 기법으로 비밀키 및 패턴 인식 기술, 정적인 단일 영상인증 기법이 적용되고 있으나 정적인 킷값의 유출 또는사진과같은영상을활용하여접근이가능하다는문제를 가지고 있다. 본 논문에서는 이러한 위험요소 및 문제로부터 스마트폰을 보호하기 위한 기술로 사용자 얼굴 인증 기술을 제안한다. 제안 기술은 사용자의 얼굴 동영상의 키 프레임을 실시간으로 추출하여 사용자를 인증하고 스마트폰의 접근을 제어한다. 인증 정보는 다수의 키 프레임으로 구성되며, 영상의 화소 및 휘도의 DC 값을 활용한 유사도 판별 알고리즘으로 사용자의 접근을 통제한다.