P.S. Tsai, and M. Shah, "A Fast Linear Shape from Shading," In Proc. Conference on Computer Vision and Pattern Recognition, Urbana/Champaign, IL, pp. 459-465, 1992  Home/Search   Document Not in Database   Summary   Related Articles   Check

....method to more than 150 face images from the Yale and Weizmann databases. Though the purpose of rendering prototype images is to improve the recognition performance, we would like to visualize the quality of the rendered images and compare them to the images obtained using a local SFS algorithm [30]. These results (Fig. 9) clearly indicate the superior quality of the prototype images rendered by our approach. More rendered prototype images using only the direct computation are plotted in Fig. 10. original) SFS) our approach) real) Figure 9: Image rendering comparison. 4.2.2 Enhancing ....

P.S. Tsai, and M. Shah, "A Fast Linear Shape from Shading," In Proc. Conference on Computer Vision and Pattern Recognition, Urbana/Champaign, IL, pp. 459-465, 1992

....in the image. These methods require a priori information and their performance depends on the accuracy of that information. Finally there are a number of methods that use local information to reconstruct depth. 30] and [22] use a rather restrictive local spherical approximation, while [31] and [39] use linear approximations of the reflectance function. In this paper we will focus on single image methods. Although photometric stereo approaches [41, 1, 12] give in general better reconstructions than single view methods, there can be no motion in the scene and in the camera position while the ....

.... we will haveanm component constraint vector C = C 1 #C 2 #: #C m ] C is a nonlinear constraint with respect to the model parameters, which in the traditional SFS formulations resulted either in nonlinear first order differential equations (PDE s) or in attempts to linearize the constraint [31, 23, 39]. PDE s require appropriate boundary conditions [17] that are often not available, whereas linearization introduces additional error. Instead, we incorporate the above brightness constraint as a nonlinear holonomic constraintin the deformable model framework, presented in a following section. ....

[Article contains additional citation context not shown here]

P.S. Tsai and M. Shah. A fast linear shape from shading. In CVPR92, pages 734--736, 1992.

....F given the hypothesized light source direction represented by ff and . One advantage of using a 3D face model is that both attached shadow and cast shadow effects can be handled. Figure 15 shows some comparisons between rendered images obtained using this method and using a local SFS algorithm [190]. Significant performance improvements have been reported when the prototype images are used in a subspace LDA system in place of the original input images (Fig. 16) 6.2 The pose problem in face recognition The performance of face recognition systems also drops significantly when pose ....

P.S. Tsai and M. Shah, "A Fast Linear Shape from Shading," in Proceedings, IEEE Conference on Computer Vision and Pattern Recognition, pp. 459-465, 1992, 65

....presented attractivity points. The scene is segmented into several depth planes to reduce the amount of hypotheses. Thus an efficient invariant object recognition is possible. For reconstruction of a scene s depth structure stereo images are used instead of shape from shading (e.g. 7] or [12]) or shape from texture methods (e.g. 4] because real images as used in the NAVIS domains include only few regions that provide sufficient data. Thus these methods can only generate depth information at a few specific points. Since for the desired applications a more detailed map is desirable, ....

Tsai P.-S., Shah M. (1992) A Fast Linear Shape from Shading. In: Proceedings of the Conference on Computer Vision and Pattern Recognition, pp. 734-736, IEEE Computer Society Press 1992

....different illuminations. We have applied our lightsource estimation and direct prototype image rendering method to more than 150 face images from the Yale and Weizmann databases. We visualize the quality of the rendered images and compare them to the images obtained using a local SFS algorithm [13]. These results (Fig. 3) clearly indicate the superior quality of the prototype images rendered by our approach. Enhancing Face Recognition In this experiment we demonstrate that the recognition rate of subspace LDA can be greatly enhanced. We conducted two independent experiments on the Yale and ....

P.S. Tsai, and M. Shah, "A Fast Linear Shape from Shading," in Proc. Conference on Computer Vision and Pattern Recognition, pp. 459-465, 1992.

....the resulting digital model. III. Shape from Shading using Perspective Projection and Camera Calibration Shape From Shading (SFS) has been primarily studied by Horn [18] and his colleagues at MIT. Since then there have been many recent developments in the algorithms [19] 20] 21] Tsai and Shah [22] have described a linear technique to solve the SFS problem that is more suitable for our dental application because of its speed and accuracy . This is important because the reconstruction of the whole jaw requires the application of the SFS algorithm on more than 30 images. However, their ....

....using perspective projection were found in the literature[23] 24] 25] In these approaches the measurements obtained are not metric as they lack the information about the camera parameters. In this section we present a SFS approach that uses the linear approximation implemented by Tsai and Shah [22] and extend it to incorporate the camera perspective projection. A camera calibration process which takes into consideration both the intrinsic and the extrinsic camera parameters is done once before acquiring the sequence of images. To calibrate the camera, we can obtain the relation between the ....

[Article contains additional citation context not shown here]

P. S. Tsai and M. Shah, "A fast linear shape from shading, " IEEEConference on Computer Vision and Pattern Recognition , pp. 734--736, July 1992.

....resembles the human head and jaw. II. Shape from Shading using Perspective Projection and Camera Calibration Shape From Shading (SFS) has been primarily studied by Horn [6] and his colleagues at MIT. Since then there have been many recent developments in the algorithms [7] 8] 9] Tsai and Shah [10] have described a linear technique to solve the SFS problem that is more suitable for our dental application because of its speed and accuracy . This is important because the reconstruction of the whole jaw requires the application of the SFS algorithm on more than 30 images. However, their ....

....using perspective projection were found in the literature[11] 12] 13] In these approaches the measurements obtained are not metric as they lack the information about the camera parameters. In this section we present a SFS approach that uses the linear approximation implemented by Tsai and Shah [10] and extend it to incorporate the camera perspective projection. A camera calibration process which takes into consideration both the intrinsic and the extrinsic camera parameters is done once before acquiring the sequence of images. To calibrate the camera, we can obtain the relation between the ....

[Article contains additional citation context not shown here]

P. S. Tsai and M. Shah, "A fast linear shape from shading, " IEEE Conference on Computer Vision and Pattern Recognition , pp. 734--736, July 1992.

....linear shape from shading algorithm (applied to the right stereo image) is shown in Figure 7. f) Since the tomato is similar to a spherical object, and it is well known that the linear shape from shading method proposed by Pentland does not compute a good depth map for spherical surfaces [16], the average gradient square error is about 1.85. The result obtained by integrating stereo and shading using our method is shown in Figure 7. d) the 16 average gradient square error reduces to 0.24. This is approximately a 48 improvement over the stereo, and a 98 improvement over shading. We ....

Tsai, Ping-Sing and Shah, Mubarak. A fast linear shape from shading. Proceedings of Computer Vision and Pattern Recognition, pp. 734-736 , 1992.

....linear shape from shading algorithm (applied to the right stereo image) is shown in Figure 7. f) Since the tomato is similar to a spherical object, and it is well known that the linear shape from shading method proposed by Pentland does not compute a good depth map for spherical surfaces [16], the average gradient square error is about 1.85. The result obtained by integrating stereo and shading using our method is shown in Figure 7. d) the 16 average gradient square error reduces to 0.24. This is approximately a 48 improvement over the stereo, and a 98 improvement over shading. ....

Tsai, Ping-Sing and Shah, Mubarak. A fast linear shape from shading. Proceedings of Computer Vision and Pattern Recognition, pp. 734-736 , 1992.

No context found.

P. S. Tsai and M. Shah, #A fast linear shape from shading," IEEE Conferenceon Computer Vision and Pattern Recognition , pp. 734#736, July 1992.

No context found.

P.-S. Tsai and M. Shah (1992), A fast linear shape from shading, Proceedings IEEE Conf. on Computer Vision and Pattern Recognition, (June 1992)734-736.

No context found.

Tsai, P.S. and Shah, M. 1992. A Fast Linear Shape from Shading. In Proc. Conference on Computer Vision and Pattern Recognition, Urbana/Champaign, IL, pp. 459-465,

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