• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권1호
• /
• pp.224-241
• /
• 2015

Multiview plus depth (MVD) videos are widely used in free-viewpoint TV systems. The best-known technique to determine depth information is based on stereo vision. In this paper, we propose a novel local stereo matching algorithm which is radiometric invariant. The key idea is to use a combined matching cost of intensity and gradient based similarity measure. In addition, we realize an adaptive cost aggregation scheme by constructing an adaptive support window for each pixel, which can solve the boundary and low texture problems. In the disparity refinement process, we propose a four-step post-processing technique to handle outliers and occlusions. Moreover, we conduct stereo reconstruction tests to verify the performance of the algorithm more intuitively. Experimental results show that the proposed method is effective and robust against local radiometric distortion. It has an average error of 5.93% on the Middlebury benchmark and is compatible to the state-of-art local methods.

• 한국정보통신학회논문지
• /
• 제16권7호
• /
• pp.1437-1446
• /
• 2012

본 논문에서는 스테레오 카메라 환경에서 고속으로 깊이 정보를 얻기 위한 응용분야에 적합한 스테레오 정합 기법을 제안하고자 한다. 이러한 조건을 만족하기 위해서 DWT 영역에서의 주파수 정보와 가변 정합창을 이용하는 적응적인 스테레오 정합 기법을 제안한다. 공간 영역에서 영상의 국부적인 특성을 분석하여 정합창의 크기를 결정하고, 주파수 영역에서 영상의 주파수 특성을 분석하여 정합창의 형태 및 스케일링 요소를 결정한다. 주파수 영역에 대한 정보를 이용하기 위해서 로컬 DWT와 전역 DWT를 활용하는 기법을 모두 적용하였다. 본 논문은 스테레오 정합을 위한 제안한 기법은 유사한 수준의 복잡도에서 고정 정합창 기반의 기법과 비교할 때 PSNR이 향상되는 것을 확인하였다.

• 대한기계학회논문집A
• /
• 제25권3호
• /
• pp.461-469
• /
• 2001

Existing area-based stereo algorithms rely heavily on rectangular windows for computing correspondence. While the algorithms with the rectangular windows are efficient, they generate relatively large matching errors due to variations of disparity profiles near depth discontinuities and doesnt take into account local deformations of the windows due to projective distortion. In this paper, in order to deal with these problems, a new correlation function with 4 directional line masks, based on robust estimator, is proposed for the selection of potential matching points. These points is selected to consider depth discontinuities and reduce effects on outliers. The proposed matching method finds an arbitrarily-shaped variable window around a pixel in the 3d array which is constructed with the selected matching points. In addition, the method take into account the local deformation of the variable window with a constant disparity, and perform the estimation of sub-pixel disparities. Experiments with various synthetic images show that the proposed technique significantly reduces matching errors both in the vicinity of depth discontinuities and in continuously smooth areas, and also does not be affected drastically due to outlier and noise.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권9호
• /
• pp.3721-3737
• /
• 2015

3D reconstruction based on binocular data is significant to machine vision. In our method, we propose a new and high efficiency 3D reconstruction approach by using a consumer camera aiming to: 1) address the configuration problem of dual camera in the binocular reconstruction system; 2) address stereo matching can hardly be done well problem in both time computing and precision. The kernel feature is firstly proposed in calibration stage to rectify the epipolar. Then, we segment the objects in the camera into background and foreground, for which system obtains the disparity by different method: local window matching and kernel feature-based matching. Extensive experiments demonstrate our proposed algorithm represents accurate 3D model.

• 한국통신학회논문지
• /
• 제32권1C호
• /
• pp.9-15
• /
• 2007

본 논문에서는 비디오 부호화의 움직임 예측에 사용되는 블록 정합 알고리즘의 계산량을 줄이는 고속 전영역 탐색 알고리즘을 제안한다. 블록 정합 알고리즘에서 사용되는 기존의 나선형 탐색 방법은 탐색 영역의 중심에서 시작하여 탐색 지점을 화소 단위로 이동하면서 움직임 예측을 수행하기 때문에 움직임이 적은 영상에 적합하다. 본 논문에서는 탐색 영역을 작은 영역으로 세분화한 후 각 영역을 새로운 탐색 순서에 따라 움직임 예측을 수행함으로써 움직임이 많은 영상을 효과적으로 탐색할 수 있다. 또한 움직임 벡터 판정시 영상의 복잡도에 따라 최적의 순서로 비용을 계산하여 복잡도를 줄이는 방법을 제안한다. 실험 결과에서 제안하는 알고리즘이 기존의 나선형 전역 탐색 방법에 비해 예측화질의 열화 없이 최대 99%까지 계산량을 감소시키는 것을 확인할 수 있다.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2000년도 추계학술대회 논문집
• /
• pp.666-669
• /
• 2000

Localization is the process of aligning the robot's local coordinates with the global coordinates of a map. A mobile robot's location is basically computed by a dead reckoning scheme, but this position information becomes increasingly inaccurate during navigation due to odometry errors. In this paper, the method of building a map of a robot's environment using ultrasonic sensor data and the occupancy grid map scheme is briefly presented. Then, the search and matching algorithms to compensate for the odometry error by comparing the local map with the reference map are proposed and verified by experiments. It is shown that the compensated error is not accumulated and exists within the limited range.

• 전자공학회논문지CI
• /
• 제46권1호
• /
• pp.136-143
• /
• 2009

일반적으로 전역적인 방법의 스테레오 매칭은 보다 높은 정확도의 변위 정보(disparity information)를 추정하는데 반해 지역적인 방법은 비교적 낮은 정확도를 가지지만 전역적인 방법보다 단순하고 빠른 방법이라고 알려져 있다. 하지만 최근 이미지의 영역 분할 방법이나 적응적 가중치를 이용한 방법 등 지역적인 방법에서 보다 정확한 결과를 얻을 수 있는 방법들이 제시되고 있다. 일반적으로 보다 빠른 스테레오 매칭을 위해 전역적인 방법보다 지역적인 방법을 주로 사용하지만 최근 제시된 지역적인 방법들은 보다 정확한 변위 정보가 추정 가능한 만큼 높은 계산적 비용이 요구되고 또한 사용되는 윈도우의 크기에 비례하여 계산 비용이 기하급수적으로 증가하는 문제점을 갖는다. 본 논문에서는 Bilateral 필터에서 사용되는 가중치와 유사하게 픽셀간 거리의 차이와 밝기 값의 차이에 따른 지역적 가중치를 이용한 스테레오 방법을 제안하고 지역적 가중치가 균등하게 정의되었을 경우 상수 시간 O(1) 내에 스테레오 매칭을 수행하는 방법을 제안한다. 본 논문에서 제안하는 방법은 실험을 통하여 기존의 우수한 방법과 비교하여 정확성 측면에서 크게 떨어지지 않으면서도 윈도우의 크기에 무관하게 상수시간의 수행 속도를 가짐을 보인다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 하계종합학술대회 논문집(4)
• /
• pp.43-46
• /
• 2000

In this paper, we propose a fast adaptive diamond search algorithm(FADS) for block matching motion estimation. Fast motion estimation algorithms reduce the computational complexity by using the UESA (Unimodal Error Search Assumption) that the matching error monotonically increases as the search moves away from the global minimum error. Recently many fast BMAs(Block Matching Algorithms) make use of the fact that the global minimum points in real world video sequences are centered at the position of zero motion. But these BMAs, especially in large motion, are easily trapped into the local minima and result in poor matching accuracy. So, we propose a new motion estimation algorithm using the spatial correlation among the adjacent blocks. We change the origin of search window according to the spatially adjacent motion vectors and their MAE(Mean Absolute Error). The computer simulation shows that the proposed algorithm has almost the same computational complexity with UCBDS(Unrestricted Center-Biased Diamond Search)〔1〕, but enhance PSNR. Moreover, the proposed algorithm gives almost the same PSNR as that of FS(Full Search), even for the large motion case, with half the computational load.

• 전자공학회논문지B
• /
• 제32B권12호
• /
• pp.1591-1602
• /
• 1995

A hybrid approach which includes edge- and region-based methods is considered. The modified non-linear Laplacian(MNL) filter is used for feature extraction. The matching algorithm has three steps which are edge, signed region, and residual region matching. At first, the edge points are matched using the sign and direction of edges. Then, the disparity is propagated from edge to inside region. A variable window is used to consider the local method which give accurate matched points and area-based method which can obtain full-resolution disparity map. In addition, a new relaxation algorithm for considering matching possibility derived from normalized error and regional continuity constraint is proposed to reduce the mismatched points. By the result of simulation for various images, this algorithm is insensitive to noise and gives full- resolution disparity map.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 하계종합학술대회 논문집(4)
• /
• pp.295-298
• /
• 2002

This paper presents a fast block-matching algorithm to improve the conventional Three-Step Search (TSS) based method. The proposed Comparison Fast Block Matching Algorithm (CFBMA) begins with DAB for adaptive search range to choose searching method, and searches a part of search window that has high possibility of motion vector like other partial search algorithms. The CFBMA also considers the opposite direction to reduce local minimum, which is ignored in almost conventional based partial search algorithms. CFBMA uses the summation half-stop technique to reduce the computational load. Experimental results show that the proposed algorithm achieves the high computational complexity compression effect and very close or better image quality compared with TSS, SES, NTSS based partial search algorithms.