Apparent 3-D camera velocity-extraction and applications

Jasinschi, Radu; Naveen, T.; Tabatabai, A.; Babib-Vovk, P.

doi:10.1109/76.875522

Cited by 3 publications

(10 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, the method discussed here [67,68] generalizes the photo-mosaic generation methods in that it: (i) computes explicitly 3-D camera parameters; (ii) determines the structure (depth) of static scenes; (iii) integrates camera and depth information with a perspective transformation model of image point displacements; (iv) models explicitly variations in image illumination; (v) can deal with arbitrary camera velocities in that it is described by an image multiresolution method; (vi) can be extended to deal with 3-D surfaces. In our approach shown in Figure 16, we identify regions or layers of "uniform" depth in the scene, and generate a 2-D sprite for each such layer.…”

Section: Layered Mosaics Based On 3-d Informationmentioning

confidence: 99%

“…The limitations of 8PSFM are its sensitivity to noise (SVD matrix computation) and the choice of points on the object. We proposed [67] a generalization of 8PSFM, called G8PSFM that avoids the need of choosing in an ad-hoc manner (image projected) object points and it makes the computation of SFM more robust.…”

Section: Structure-from-motion: Generalized Eight-point Algorithmmentioning

confidence: 99%

“…We devised a robust sub-optimum statistical approach that determines the choice of the 'best' set of camera motion parameters from a subset of all possible combinations. This uses an overall velocity estimation error as the quality measure (see [67]). …”

Section: Structure-from-motion: Generalized Eight-point Algorithmmentioning

confidence: 99%

“…The first is called Generative Video [62][63][64][65] and the second one [66][67][68] uses partial or full 3-D information to generate mosaics. The motivation in these two approaches is to overcome the limitations of classical mosaic generation methods: the objects in the (3-D) scene are very far from the camera so that there barely exists any parallax 1 .…”

Section: Mosaics: From 2-d To 3-dmentioning

confidence: 99%

“…This generalizes GV by using this extra 3-D information for mosaic generation. At the core of this method [66][67][68] lies the detailed extraction of 3-D information, i.e., structure-from-motion.…”

Section: -D Based Mosaicsmentioning

confidence: 99%

See 4 more Smart Citations

Content-based image sequence representation

Aguiar¹,

Jasinschi²,

Moura³

et al. 2004

Computer Engineering Series

View full text Add to dashboard Cite

Abstract. In this chapter we overview methods that represent video sequences in terms of their content. These methods differ from those developed for MPEG/H.26X coding standards in that sequences are described in terms of extended images instead of collections of frames. We describe how these extended images, e.g., mosaics, are generated by the basically same principle: the incremental composition of visual photometric, geometric, and multi-view information into one or more extended images. Different outputs, e.g., from single 2-D mosaics to full 3-D mosaics, are obtained depending on the quality and quantity of photometric, geometric, and multi-view information. In particular, we detail a framework well suited to the representation of scenes with independently moving objects. We address the two following important cases: i) the moving objects can be represented by 2-D silhouettes (generative video approach) ; or ii) the camera motion is such that the moving object must be described by their 3-D shape (recovered through rank 1 surface-based factorization). A basic pre-processing step in content-based image sequence representation is to extract and track the relevant background and foreground objects. This is achieved by 2-D shape segmentation for which there is a wealth of methods and approaches. The chapter includes a brief description of active contour methods for image segmentation.

show abstract

Section: Layered Mosaics Based On 3-d Informationmentioning

confidence: 99%

Section: Structure-from-motion: Generalized Eight-point Algorithmmentioning

confidence: 99%

Section: Structure-from-motion: Generalized Eight-point Algorithmmentioning

confidence: 99%