Traditional plane alignment techniques are typically performed between pairs of frames. In this paper we present a method for extending existing two-frame planar-motion estimation techniques into a simultaneous multi-...
详细信息
Traditional plane alignment techniques are typically performed between pairs of frames. In this paper we present a method for extending existing two-frame planar-motion estimation techniques into a simultaneous multi-frame estimation, by exploiting multi-frame geometric constraints of planar surfaces. The paper has three main contributions: (i) we show that when the camera calibration does not change, the collection of all parametric image motions of a planar surface in the scene across multiple frames is embedded in a low dimensional linear subspace; (ii) we show that the relative image motion of multiple planar surfaces across multiple frames is embedded in a yet lower dimensional linear subspace, even with varying camera calibration; and (iii) we show how these multi-frame constraints can be incorporated into simultaneous multi-frame estimation of planar motion, without explicitly recovering any 3D information, or camera calibration. The resulting multi-frame estimation process is more constrained than the individual two-frame estimations, leading to more accurate alignment, even when applied to small image regions.
We consider a recent approach to recognition that uses regions to determine the pose of objects while allowing for partial occlusion of the regions. We further analyze properties of the method for planar objects under...
详细信息
We consider a recent approach to recognition that uses regions to determine the pose of objects while allowing for partial occlusion of the regions. We further analyze properties of the method for planar objects undergoing projective transformations. We prove that three visible regions are sufficient to determine the transformation uniquely, and that for a large class of objects two regions are insufficient. However, we show that when several regions are available, the pose of the object can generally be recovered even when all but two regions are significantly occluded. Our analysis is based on investigating the flow patterns of points under projective transformations in the presence of fixed points.
A new way of segmenting an image based on visual importance is proposed. The image is partitioned into blocks with variable sizes according to the visual information inside the block. Variably sized vector quantizatio...
详细信息
A new way of segmenting an image based on visual importance is proposed. The image is partitioned into blocks with variable sizes according to the visual information inside the block. Variably sized vector quantization and subband coding are then applied to encode the image. The proposed procedure shows a significant reduction in bit rate usage without the loss of subjective and objective quality of images.
The Generalized Lambda Distribution, GLD(λ1, λ2, λ3, λ4), is a four-parameter family that has been used for fitting distributions to data sets. With rare exceptions, the method of moments has been used for estimat...
详细信息
The Generalized Lambda Distribution, GLD(λ1, λ2, λ3, λ4), is a four-parameter family that has been used for fitting distributions to data sets. With rare exceptions, the method of moments has been used for estimating the GLD parameters, making it impossible to fit a GLD distribution whose first four moments do not exist. This paper deals with the use of an estimation method that is based on percentiles. To provide practitioners ready access to this method, extensive tables are provided to facilitate parameter estimation.
We describe an error in earlier probabilistic analyses of the pure literal heuristic as a procedure for solving the satisfiability problem for sets of k-clauses (k-SAT). All probabilistic analyses are in the constant ...
详细信息
We describe an error in earlier probabilistic analyses of the pure literal heuristic as a procedure for solving the satisfiability problem for sets of k-clauses (k-SAT). All probabilistic analyses are in the constant degree model in which a random instance C of k-SAT consists of m clauses selected independently and uniformly (with replacement) from the set of all k-clauses over n variables. We provide a new analysis for k = 2. Specifically, we show with probability approaching 1 as m goes to ∞ one can apply the pure literal rule repeatedly to a random instance of 2-SAT until the number of clauses is `small' provided n/m≥λ>1. But if n/m&leλΕ clauses will remain.
In this paper; we show how to extend the argument due to Bonet, Pitassi and Raz to show that bounded-depth Frege proofs do not have feasible interpolation, assuming that factoring of Blum integers or computing the Dif...
详细信息
In this paper; we show how to extend the argument due to Bonet, Pitassi and Raz to show that bounded-depth Frege proofs do not have feasible interpolation, assuming that factoring of Blum integers or computing the Diffie-Hellman function is sufficiently hard. It follows as a corollary that bounded-depth Frege is not automatizable; in other words, there is no deterministic polynomial-time algorithm that will output a short proof if one exists. A notable feature of our argument is its simplicity.
In this article, we discuss the application, of a Green's function approach to contact maps. We apply the contact map algorithm to the case of an elastic bar impacting on a massive foundation. It is demonstrated t...
详细信息
In this article, we discuss the application, of a Green's function approach to contact maps. We apply the contact map algorithm to the case of an elastic bar impacting on a massive foundation. It is demonstrated that the energy lost to elastic waves is an important factor in post-contact states.
暂无评论