We present a method to compute depth from the amount of defocus in two images obtained from the same view-point but with different camera parameter settings. The change in defocus (blur) between the two images is prop...
详细信息
We present a method to compute depth from the amount of defocus in two images obtained from the same view-point but with different camera parameter settings. The change in defocus (blur) between the two images is proportional to the depth in the scene. We introduce a novel method to estimate the blur using a multiresolution local frequency representation of the input image pair. A confidence measure is used to discriminate between high error and low error blur estimates. Quantitative experimental results are shown for both real and synthetic images.
Optical flow computation may be divided into four processing steps where the first is extraction of image features suitable for flow estimation. Using a generalization of the basic flow constraint it is possible to es...
详细信息
Optical flow computation may be divided into four processing steps where the first is extraction of image features suitable for flow estimation. Using a generalization of the basic flow constraint it is possible to estimate flow vectors from a set of feature images obtained from the input image sequence. Generalized Sobel operators provide a suitable set of feature extraction operators. The set of five filters up to second order provides good approximations of ideal edge and line detectors. A review of spatial and frequency constraints on optical flow suggests a multiresolution approach to optical flow where the initial feature extractors consists of velocity tuned operators. We show that two-dimensional generalized Sobel operators may be extended to spatio-temporal velocity tuned filters for optical flow estimation. Experimental results compare well to existing methods.< >
In this paper, we present the design and VLSI implementation of a finite impulse response (FIR) filter using silicon compilation tools. The particular silicon compilation tools we used was Olympus, which is public-dom...
详细信息
In this paper, we present the design and VLSI implementation of a finite impulse response (FIR) filter using silicon compilation tools. The particular silicon compilation tools we used was Olympus, which is public-domain set of vertically integrated CAD tools for VLSI design developed at Stanford University. The design was based on using the distributed arithmetic method. Our results are very encouraging and this project tests the Olympus set of tools for a digital signalprocessing chip design application. We also present our finding about this set of published tools which has not been used for this kind of design problems before now.
In this paper, we present a comprehensive theory of spatially-variant (SV) mathematical morphology. A kernel representation of increasing operators in terms of the union (resp., intersection) of SV erosions (resp., SV...
详细信息
In this paper, we present a comprehensive theory of spatially-variant (SV) mathematical morphology. A kernel representation of increasing operators in terms of the union (resp., intersection) of SV erosions (resp., SV dilations) is provided. A representation of algebraic openings (resp., algebraic closings) in terms of the union (resp., intersection) of SV openings (resp., SV closings) is also provided.< >
Singular behavior of PD-eigenvalues of an nth-order scalar polynomial differential operator (SPDO) /spl Dscrsub /spl alpha=/spl deltasup n/+/spl Sigmasub k=1sup nspl alphasub k/(t)/spl deltasup k/spl minus/1/, where /...
详细信息
Singular behavior of PD-eigenvalues of an nth-order scalar polynomial differential operator (SPDO) /spl Dscrsub /spl alpha=/spl deltasup n/+/spl Sigmasub k=1sup nspl alphasub k/(t)/spl deltasup k/spl minus/1/, where /spl delta/=d/dt, is investigated. Main results of this paper include: (i) definition and properties of PD-eigenvectors associated with PD-eigenvalues, (ii) definitions and properties of generalized PD-eigenvectors and generalized PD-eigenvalues for singular PD-eigenvalues; (iii) application of (i) and (ii) in stability analysis of linear time-varying (LTV) systems /spl Dscrsub /spl alpha/spl lcub/y/spl rcub/=0, and (iv) application (i) and (ii) in the realization of PD-characteristic equation. The new results will have a significant impact on applications of the unified eigenvalue theory to the analysis and control of LTV control systems, and its further development.< >
This study constitutes an effort in identifying a class of unbounded coordinate transformations for vector linear time-varying systems of the form x/spl dot/=A(t)x. Main results of this paper include: (i) the definiti...
详细信息
This study constitutes an effort in identifying a class of unbounded coordinate transformations for vector linear time-varying systems of the form x/spl dot/=A(t)x. Main results of this paper include: (i) the definition and properties of meromorphic D-similarity transformations, (ii) two theorems on stability preserving meromorphic transformations. The results appear to be somewhat restrictive, but certainly shed some light on this intriguing yet difficult problem.< >
A novel class of Learning Vector Quantizers (LVQs) based on multivariate order statistics is proposed in order to overcome the drawback that the estimators for obtaining the reference vectors in LVQ do not have robust...
详细信息
A novel class of Learning Vector Quantizers (LVQs) based on multivariate order statistics is proposed in order to overcome the drawback that the estimators for obtaining the reference vectors in LVQ do not have robustness either against erroneous choices for the winner vector or against the outliers that may exist in vector-valued observations. The performance of the proposed variants of LVQ is demonstrated by experiments. In the case of marginal median LVQ, its asymptotic properties are derived as well.< >
This paper analyzes the problem of global asymptotic stability of delta-operator formulated discrete-time systems implemented in fixed-point arithmetic. It is shown that the free response of such a system tends to pro...
详细信息
This paper analyzes the problem of global asymptotic stability of delta-operator formulated discrete-time systems implemented in fixed-point arithmetic. It is shown that the free response of such a system tends to produce period one limit cycles if conventional quantization arithmetic schemes are used. Explicit necessary conditions for global asymptotic stability are derived, and these demonstrate that, in almost all cases, fixed-point arithmetic does not allow for global asymptotic stability in delta-operator formulated discrete-time systems that use a short sampling time.< >
The convergence properties of linearly stable multi-dimensional systems are investigated for the case of delta-operator implementations in fixed-point format. It is shown that zero-convergence is almost never achieved...
详细信息
The convergence properties of linearly stable multi-dimensional systems are investigated for the case of delta-operator implementations in fixed-point format. It is shown that zero-convergence is almost never achieved, if the sampling time is small. Using a one-dimensional analysis, it is demonstrated that zero-convergence cannot be guaranteed along the axis of the first hyper-quadrant for a first hyper-quadrant causal system. This limits the use of delta-operators for solving partial differential equations in discrete time with fixed-point arithmetic.< >
A prototype of an optical-electronic histogram generator has been designed and tested for 1-D objects. In this scheme, the object to be analyzed is laser scanned. The resulting optical signal is detected by a photodet...
详细信息
A prototype of an optical-electronic histogram generator has been designed and tested for 1-D objects. In this scheme, the object to be analyzed is laser scanned. The resulting optical signal is detected by a photodetector which generates an electricalsignal output which is subsequently analyzed with a combination of analog and digital electronics. The system has shown to be fairly modular in design. Various aspects of the extension of the design to two dimensions is discussed.< >
暂无评论