• Journal of Information Processing Systems
• /
• 제13권2호
• /
• pp.256-267
• /
• 2017

An active research area in computer vision, stereo matching is aimed at obtaining three-dimensional (3D) information from a stereo image pair captured by a stereo camera. To extract accurate 3D information, a number of studies have examined stereo matching algorithms that employ adaptive support weight. Among them, the adaptive census transform (ACT) algorithm has yielded a relatively strong matching capability. The drawbacks of the ACT, however, are that it produces low matching accuracy at the border of an object and is vulnerable to noise. To mitigate these drawbacks, this paper proposes and analyzes the features of an improved stereo matching algorithm that not only enhances matching accuracy but also is also robust to noise. The proposed algorithm, based on the ACT, adopts the truncated absolute difference and the multiple sparse windows method. The experimental results show that compared to the ACT, the proposed algorithm reduces the average error rate of depth maps on Middlebury dataset images by as much as 2% and that is has a strong robustness to noise.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1987년도 한국자동제어학술회의논문집; 한국과학기술대학, 충남; 16-17 Oct. 1987
• /
• pp.96-99
• /
• 1987

In the way of perception and depth finding of 3-D scene stereo vision is one of the best ways. Matching is important problem of stereo vision. In this paper we develop an algorithm that is adaptable to errors and computation time. And we demonstrate the way of matching.

• 대한원격탐사학회지
• /
• 제18권3호
• /
• pp.181-186
• /
• 2002

One of the most important procedure in stereo DEM generation is the stereo matching process which finds the conjugate pixels in a pair of stereo imagery. In order to be found as conjugate pixels, the pixels should have distinct spatial feature to be distinguished from other pixels. However, in the homogeneous areas such as water covered or forest canopied areas, it is very difficult to find the conjugate pixels due to the lack of distinct spatial feature. Most of erroneous elevation values in the stereo DEM are produced in those homogeneous areas. This paper presents a simple method for improving the quality of stereo DEM utilizing the texture filters. An entropy filter was applied to one of the input stereo imagery to extract very homogeneous areas before stereo matching process. Those extracted homogeneous areas were excluded from being candidates for stereo matching process. Also a statistical texture filter was applied to the generated elevation values before the interpolation process was applied in odor to remove the remaining anomalous elevation values. Stereo pair of SPOT level 1B panchromatic imagery were used for the experiments. The results showed that by utilizing the texture filters as a pre and a post processor of stereo matching process, the quality of the stereo DEM could be dramatically improved.

• 전자공학회논문지SC
• /
• 제37권2호
• /
• pp.50-59
• /
• 2000

스테레오 영상(stereo image)이란 같은 물체나 장면을 담고있는 서로 다른 시점의 두 영상이며 스테레오 영상에서 깊이 정보를 얻어내는 것을 스테레오 비전(stereo vision)이라 한다. 스테레오 비전에서 가장 중요한 과정은 두 영상에서 서로 일치하는 점을 찾아내는 스테레오 정합(stereo matching)이다. 그러나, 스테레오 정합은 매우 많은 계산을 필요로 하기 때문에 실시간으로 정합하기 어렵다. 본 논문에서는 실시간으로 스테레오 정합을 처리할 수 있는 스테레오 영상 정합 프로세서(stereo image match procesor：SIMP)를 설계하고 구현하였다. 이를 위해 슬라이딩 메모리(sliding memory)와 최소 선택 트리(minimum selection tree)를 제안하였고 파이프라인 구조(pipeline architecture)와 병렬 처리 기법을 이용하였다. SIMP의 입력은 64 그레이 레벨인 두 개의 64×64 스테레오 영상이고 출력은 최대 7의 값을 가지는 변이(disparity)와 12비트의 주소로 이들을 이용하여 64×64 변이도(disparity map)를 구성할 수 있다. SIMP는 약 240 프레임/초의 속도로 스테레오 영상을 처리할 수 있다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 추계학술대회 논문집 학회본부 D
• /
• pp.747-749
• /
• 2000

In this paper, we proposed color stereo matching algorithm using dynamic programming. The conventional gray stereo matching algorithms show blur at depth discontinuities and non-existence of matching pixel in occlusion lesions. Also it accompanies matching error by lack of matching information in the untextured region. This paper defines new cost function makes up for the problems happening in conventional gray stereo matching algorithm. New cost function contain the following properties. I) Edge points are corresponded to edge points. ii) Non-edge points are corresponded to non-edge points. iii) In case of exiting the amount of edges, the cost function has some weight in proportion to path distance. Proposed algorithm was applied in various images obtained by parallel camera model. As the result, proposed algorithm showed improved performance in the aspect of matching error and processing in the occlusion region compared to conventional gray stereo matching algorithms.

• 한국통신학회논문지
• /
• 제35권8C호
• /
• pp.680-688
• /
• 2010

본 논문에서는 영상 분할(image segmentation)을 이용한 다이내믹 프로그래밍(dynamic programming, DP) 기반의 스테레오 정합 (stereo matching)기법을 제안한다. 다이내믹 프로그래밍은 스테레오 정합을 포함하는 여러 가지 컴퓨터 비젼 문제들의 고전적이고 인기가 있는 최적화 방법이다. 그러나 스테레오 정합 문제에 적용될 때 스캔라인들 사이의 수직 상호 관계가 적절하게 고려되지 않기 때문에 기존의 DP의 성능은 만족스럽지 않다. 본 논문에서 제안하는 알고리즘에서는 영상을 분할하여 정확한 경계정보를 획득한 다음 경계 정보에 의거하여 변이의 불연속과 폐색영역을 고려한다. Middlebury 스테레오 영상에 적용한 실험 결과들은 제안된 알고리즘이 이전의 다이내믹 기반 알고리즘보다 더 좋은 성능을 보여주는 것을 입증해준다.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제53권4호
• /
• pp.277-285
• /
• 2004

We present a new hybrid stereo matching technique in terms of the co-operation of area-based stereo and feature-based stereo. The core of our technique is that feature matching is carried out by the reference of the disparity evaluated by area-based stereo. Since the reference of the disparity can significantly reduce the number of feature matching combinations, feature matching error can be drastically minimized. One requirement of the disparity to be referenced is that it should be reliable to be used in feature matching. To measure the reliability of the disparity, in this paper, we employ the self-consistency of the disunity Our suggested technique is applied to the detection of 3D line segments by 2D line matching using our hybrid stereo matching, which can be efficiently utilized in the generation of the rooftop model from urban imagery. We carry out the experiments on our hybrid stereo matching scheme. We generate synthetic images by photo-realistic simulation on Avenches data set of Ascona aerial images. Experimental results indicate that the extracted 3D line segments have an average error of 0.5m and verify our proposed scheme. In order to apply our method to the generation of 3D model in urban imagery, we carry out Preliminary experiments for rooftop generation. Since occlusions are occurred around the outlines of buildings, we experimentally suggested multi-image hybrid stereo system, based on the fusion of 3D line segments. In terms of the simple domain-specific 3D grouping scheme, we notice that an accurate 3D rooftop model can be generated. In this context, we expect that an extended 3D grouping scheme using our hybrid technique can be efficiently applied to the construction of 3D models with more general types of building rooftops.

• 대한원격탐사학회지
• /
• 제8권1호
• /
• pp.15-26
• /
• 1992

This paper describes an implementation of adaptive stereo matching for DBM generation. The matching method of two stereo satellite images to find corresponding points used in this paper is area-based matching, which is usually used in the field of making DBM. Same window size and search area used as in the conventional matching methods and we propose adaptive stereo matching algorithm in this paper. We cluster three areas which are consist of mountainous areas, cultivated areas and cities, and rivers and lakes by using proposed linear feature extracting method. These classified areas are matched by adaptive window size and search area, but rivers and lakes is excluded in this experiment. The matching time is three times faster than conventional methods.

• 정보처리학회논문지B
• /
• 제19B권2호
• /
• pp.113-122
• /
• 2012

스테레오 정합(stereo matching) 기술은 주어진 두 영상에서 동일한 물체의 영상점이 어떤 위치 관계를 가지고 있는지를 결정하는 기술이다. 본 논문에서는 영상 특징점에 대해 스테레오 위치관계를 결정하는 새로운 스테레오 특징점 정합(stereo feature matching) 방법을 제시한다. 제안하는 방법은 주어진 스테레오 영상에서 FAST 추출기를 이용하여 특징점을 추출하고, 특징점 벡터들의 정보들을 내부에 포함하는 특징창(feature window)이라는 공간을 정의하여 스테레오 정합의 성능을 향상한다. 제안하는 방법은 표준 영상에 추출된 특징점들에 대해 특징창을 생성하고, 참조 영상에서 표준 영상의 특징창과 가장 유사한 특징창을 탐색 및 결정한 다음, 결정된 두 개의 특징창 내부의 특징점들의 시차관계는 특징링크(feature link)를 생성하여 시차를 결정한다. 만약, 이 과정에서 시차가 결정되지 않은 특징점들이 있다면, 특징창 내의 결정된 시차 정보를 이용하여 시차 값을 보간한다. 마지막으로, 제안하는 방법의 성능을 검증하기 위해 결과 영상과 정답 영상의 시차를 비교하여 정합 정확성과 수행시간을 비교하였다. 또한, 기존의 특징점 기반 스테레오 정합 방법들과 제안하는 방법의 성능을 비교 및 분석하였다.

• 한국자동차공학회논문집
• /
• 제11권1호
• /
• pp.201-210
• /
• 2003

In this paper, we propose a new binocular stereo correspondence method by maximizing a fitness formulated by integrating two constraints of edge similarity and disparity smoothness simultaneously. The proposed stereopsis focusing to measure distances to leading vehicles on roads uses intensity gradients as matching attribute. In contrast to the previous work of area-based stereo matching, in which matching unit is a pixel, the matching unit of the proposed method becomes an area itself which is obtained by selecting a series of pixels enclosed by two pixels on the left and right boundaries of an object. This approach allows us to cope with real-time processing and to avoid window size selection problems arising from conventional area-based stereo.