A fast area-based stereo matching algorithm

Stefano, Luigi Di; Marchionni, M.; Mattoccia, Stefano

doi:10.1016/j.imavis.2004.03.009

Cited by 229 publications

(138 citation statements)

References 17 publications

Supporting

Mentioning

135

Contrasting

Unclassified

Order By: Relevance

“…For the best of authors' knowledge, we propose the first RGB-D benchmark dataset with handlabeled ground truth that includes sequences with different challenging scenarios for FgjBg segmentation. Moreover, we tested the proposed algorithm on the stereo data presented in [11] for which we provided a ground truth.…”

Section: Benchmark Data and Resultsmentioning

confidence: 99%

“…The proposed dataset is composed by five sequences of indoor environments, obtained with a static device that registers different demanding situations such as cast shadows, color and depth camouflage and moved background object. Moreover the dataset contains a stereo sequence presented in [11] for which we provide the ground truth. The benchmark database has been used to test the proposed strategy and state of the art algorithms.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Background foreground segmentation with RGB-D Kinect data: An efficient combination of classifiers

Camplani

Salgado

2014

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

Low cost RGB-D cameras such as the Microsoft's Kinect or the Asus's Xtion Pro are completely changing the computer vision world, as they are being successfully used in several applications and research areas. Depth data are particularly attractive and suitable for applications based on moving objects detection through foreground/background segmentation approaches; the RGB-D applications proposed in literature employ, in general, state of the art foreground/background segmentation techniques based on the depth information without taking into account the color information. The novel approach that we propose is based on a combination of classifiers that allows improving background subtraction accuracy with respect to state of the art algorithms by jointly considering color and depth data. In particular, the combination of classifiers is based on a weighted average that allows to adaptively modifying the support of each classifier in the ensemble by considering foreground detections in the previous frames and the depth and color edges. In this way, it is possible to reduce false detections due to critical issues that can not be tackled by the individual classifiers such as: shadows and illumination changes, color and depth camouflage, moved background objects and noisy depth measurements. Moreover, we propose, for the best of the author's knowledge, the first publicly available RGB-D benchmark dataset with hand-labeled ground truth of several challenging scenarios to test background/foreground segmentation algorithms.

show abstract

Section: Benchmark Data and Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Background foreground segmentation with RGB-D Kinect data: An efficient combination of classifiers

Camplani

Salgado

2014

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

show abstract

“…After obtaining the camera calibration parameters by observing different orientations of a 2D planar calibration pattern [1], the main task is to determine correspondences between the stereo images along the epipolar lines. Methods to determine such correspondence in stereo vision can be divided into two classes: global and local methods [2,3], The global method relies on an iterative scheme and obtains the disparity on the basis of the minimization of a global cost function [4,5]. This method can produce an accurate and dense disparity map, but at a high computational cost as the local method is based on the relation between each pixel and its adjacent pixels.…”

Section: Introductionmentioning

confidence: 99%

A Point Cloud Alignment Algorithm Based on Stereo Vision Using Random Pattern Projection

Chen

Huang

Yeh³

et al. 2016

AUSMT

View full text Add to dashboard Cite

This paper proposes a point cloud alignment algorithm based on stereo vision using Random Pattern Projection (RPP). In the application of stereo vision, it is rather difficult to find correspondences between stereo images of texture-less objects. To overcome this issue, RPP is used to enhance the object's features, thus increasing the accuracy of the identified correspondences of the stereo images. In the 3D alignment algorithm, the down sample technique is used to filter out the outliers of the point cloud data to improve system efficiency. Furthermore, the extracted features of the down sample point cloud data were applied in the matching process. Finally, the object's pose was estimated by the alignment algorithm based on object features. In experiments, the maximum error and standard deviation of rotation are respectively about 0.031°and 0.199°, while the maximum error and standard deviation of translation are respectively about 0.565 mm and 0.902 mm . The execution time for pose estimation is about 230ms.

show abstract

“…This constraint may be violated at depth discontinuities in the scene. Three broad classes of techniques have been used for stereo matching: area-based (Di Stefano et al, 2004;Scharstein & Szelinski, 2002), feature-based (Venkateswar & Chellappa, 1995;Dhond & Aggarwal, 1989), and phase-based (Fleet et al, 1991;Fleet, 1994). Area-based algorithms use local pixel intensities as a distance measure and they produce dense disparity maps, i.e.…”

Section: Introductionmentioning

confidence: 99%