M. Bertero, T. Poggio, and V. Torre. Ill-posed problems in early vision. Proceedings of the IEEE, 76(8):869--889, 1988.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....As such, computer vision systems ideally would like to be able to reconstruct objects or environments from a sequence of pictures. This measurement problem is inherently ill posed as projected image intensity fails to provide an invertible encoding of surface characteristics under most conditions [10]. The conditioning of the system can be described by dividing it into parts. In general, imagebased surface reconstruction system can be broken up into three elements (Figure 1.1) i) Image correspondence, ii) Depth estimation from triangulation or back projection, and iii) Depth integration. ....

Bertero, M., Poggio, T.A., and Torre, V., Ill-Posed Problems in Early Vision, Proceedings of the IEEE, Vol. 76, No. 8, pp. 869-889, August 1988.

....an assumption of integrability of the square of the second derivative of the estimand. Most nonparametric regression literature has been concerned with estimating scalar functions from linear, scalar measurements. The natural generalization of these penalized likelihood [45] or regularized [46] methods to our nonlinear, multi dimensional, object estimation problem is the following estimator and optimality criterion : #(#) #(#) #y s(# )# Zk k (z) dz, 9) This criterion is nonparametric in the sense that we have avoided using a parametric (e.g. ....

....the nonlinearity of (12) limits how much we can say about its theoretical properties. There are probably local minima, and even the global minimum is not unique in general, due to the non uniqueness discussed in Section IV. However, regularization methods have shown promise in other applications [46], and the empirical results of Section VII likewise are encouraging. We have defined an optimality criterion for the object reconstruction problem. This criterion can be used to compare suboptimal algorithms, or minimized to generate an arterial tree estimate. In the next section, we present an ....

M. Bertero, T. Poggio, and V. Torre, "Ill posed problems in early vision," Proc. IEEE, vol. 76, pp. 869--89, Aug. 1988.

....available, the link measurements, remains approximately constant. One can see the difficulty for large N the problem becomes massively underconstrained. There is extensive experience with ill posed linear inverse problems from fields as diverse as seismology, astronomy, and medical imaging [1, 2, 17, 18, 26], all leading to the conclusion that some sort of side information must be brought in, producing a result which may be good or bad depending on the quality of this information. All of the previous work on IP traffic matrix estimation has incorporated prior information: for instance, Vardi [24] and ....

....perturbations of the observations. In our case, y are the SNMP link measurements, x is the traffic matrix written as a vector, and A is the routing matrix. There is extensive experience with ill posed linear inverse problems from fields as diverse as seismology, astronomy, and medical imaging [1, 2, 17, 18, 26], all leading to the conclusion that some sort of side information must be brought in, producing a reconstruction which may be good or bad depending on the quality of the prior information. Many such proposals solve the minimization problem 2 # J(x) 10) where where # #2 denotes the L2 ....

M. Bertero, T. Poggio, and V. Torre. Ill-posed problems in early vision. In Proc. of the IEEE, 76:869--889, 1988.

....was mainly due to the better results obtained in the computation of the path direction. When contrast material decreases, failure to localize some points of the contour occurs. In order to deal with this kind of error additional constraints must be added to regularize the localization procedure [4]. 1 4 7 10 13 16 19 22 25 28 Figure 3. Results of the automatic contour tracking procedure over a sequence of 28 ventriculographics images. ....

M.Bertero, T.A.Poggio, Ill-Posed Problems in Early Vision, Proceeding of the IEEE, vol. 76, n. 8, August 1988.

....slightly outperforms MFA. Furthermore, ST allows implementation based on integer arithmetic. 1 Introduction Reconstruction of visual surfaces representing some property of a scene (brightness, shape, distance, motion, from observed image data, is known to be an ill posed inverse problem [2], 11] 17] Both reg ularization and Bayesian approaches provide ways to achieve stability by incorporating a priori knowledge or constraints into the problem. In earlier work, the solutions were restricted to some class of coutinuous and smooth functions [1] 2] 3] 17] this being ....

....an ill posed inverse problem [2] 11] 17] Both reg ularization and Bayesian approaches provide ways to achieve stability by incorporating a priori knowledge or constraints into the problem. In earlier work, the solutions were restricted to some class of coutinuous and smooth functions [1] [2], 3] 17] this being clearly unreasonable iu the pres ence of discontinuities which are key features of visual perception. The incorporation of discontinuity detection into the surface reconstruction process has been studied by several authors. Some fundamental refer ences are [4] 5] 131, ....

M. Bertero, T. Poggio, and V. Torre. "Ill-posed problems in early vision". Proceedings of the IEEE, vol. 76(8):869-889, August 1988.

.... Regularized Least Squares Image Recovery Regularization theory determines a solution by trying to include prior knowledge about the signal properties, and substitutes the ill posed problem by a well posed one whose solution is a sucient approximation to the solution of the given ill posed problem [4, 40]. A solution which exchanges delity to the data for the adherence to the constraints that represent the prior knowledge about the solution that is introduced, which in most cases is the smoothness property of the image, is pursued according to the regularization approach. The only di erence ....

M. Bertero, T. Poggi, and V. Torre, \Ill-posed problems in early vision," Proceedings of the IEEE,vol. 76, no. 8, pp. 869-889, August 1988.

No context found.

M. Bertero, T. Poggio, and V. Torre. Ill-posed problems in early vision. Proceedings of the IEEE, 76:869--889, 1988.

No context found.

M. Bertero, T. Poggio, and V. Torre. Ill-posed problems in early vision. Proceedings of the IEEE, 76:869-889, 1988.

....choice for regression. Tikhonov and Arsenin [3] and Schonberg [11] used least squares regularization to restore well posedness to ill posed regression problems. In 1988, Bertero, Poggio and Torre introduced regularization in computer vision, making use of Reproducing Kernel Hilbert Space ideas [12]. In 1989, Girosi and Poggio [13, 14] introduced classification and regression techniques with the square loss in the field of supervised learning. They used pseudodi#erential operators as their stabilizers; these are essentially equivalent to using the norm in an RKHS. In 1990, Wahba [4] ....

M. Bertero, T. Poggio and V. Torre, Ill-posed problems in early vision, Proceedings of the IEEE 76 (1988) 869--889.

....Gaussian Radial Basis Functions were proposed as an alternative to neural networks by Broomhead and Loewe. Of course, RKHS had been pioneered by Parzen and Wahba ( 37, 53] for applications closely related to learning, including data smoothing (for image processing and computer vision, see [4, 42]) A Bayesian interpretation The learning algorithm Equation 4 has an interesting Bayesian interpretation [52, 53] the data term that is the first term with the quadratic loss function is a model of (Gaussian, additive) noise and the RKHS norm (the stabilizer) corresponds to a prior ....

M. Bertero, T. Poggio, and V. Torre. Ill-posed problems in early vision. Proceedings of the IEEE, 76:869--889, 1988.

....choice for regression. Tikhonov and Arsenin [3] and Schonberg [11] used least squares regularization to restore well posedness to ill posed regression problems. In 1988, Bertero, Poggio and Torre introduced regularization in computer vision, making use of Reproducing Kernel Hilbert Space ideas [12]. In 1989, Girosi and Poggio [13, 14] introduced classification and regression techniques with the square loss in the field of supervised learning. They used pseudodi#erential operators as their stabilizers; these are essentially equivalent to using the norm in an RKHS. In 1990, Wahba [4] ....

M. Bertero, T. Poggio and V. Torre, Ill-posed problems in early vision, Proceedings of the IEEE 76 (1988) 869--889.

No context found.

M. Bertero, T. Poggio, and V. Torre. Ill-posed problems in early vision. Proceedings of the IEEE, 76(8):869--889, 1988.

No context found.

M. Bertero, T. Poggio, and V. Torre., "Ill-posed problems in early vision.," Proc. of the IEEE, vol. 76, no. 869--889, 1988.

No context found.

M. Bertero, T. Poggio, V. Torre, Ill-posed problems in early vision, Proc. IEEE 76 (August 1988) 869---889.

No context found.

M. Bertero, T.A Poggio, and V. Torre, "Ill-posed problems in early vision," in Proceedings of the IEEE, Vol. 8, pp. 869--889, 1988.

No context found.

M. Bertero, T. Poggio, and V. Torre, "Ill-posed problems in early vision". Proc. of the IEEE, Vol. 76, No. 8, pp. 869-889, August 1988.

No context found.

M. A. Bertero, T. Poggio, and V. Torre, "Ill-posed problems in early vision," Proceedings of IEEE 76(8), pp. 869--889, 1988.

No context found.

M. Bertero, T. A. Poggio, and V. Torre. Ill-posed problems in early vision. Proc. IEEE, 76(8):869--889, August 1988.

No context found.

M. Bertero, T. Poggio, and V. Torre, "Ill-Posed Problems in Early Vision," Proc. IEEE, vol. 76, pp. 869-889, 1988.

No context found.

M. Bertero, T. Poggio, and V. Torre, "Ill-posed problems in early vision," Proc. IEEE, vol. 76, pp. 869--889, 1988.

No context found.

M. Bertero, T. Poggio and V. Torre. "Ill-posed problems in early vision," Proc. IEEE, Vol. 76, pp. 869-889, 1988.

No context found.

M. Bertero, T. A. Poggio, and V. Torre. Ill-posed problems in early vision. Proc. IEEE, 76(8):869--889, August 1988.

No context found.

Bertero, M., Poggio, T.A., and Torre, V.: Ill-Posed Problems in Early Vision. Proceedings of the IEEE, Vol. 76, No. 8, pp. 869-889, August 1988.

No context found.

M. Bertero, T. Poggio, and V. Torre.Ill-posed problems in early vision. Proceedings of the IEEE, 76:869--889, 1988.

No context found.

M. Bertero, T.A. Poggio, and V. Torre, "Ill-posed problems in early vision," Proc. IEEE, vol. 76, pp. 869-889, Aug. 1988.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC