Y. Altunbasak, A.M. Tekalp, "Occlusion-adaptive, content-based mesh design and forward tracking," IEEE Trans. on Image Processing, vol. 6, no. 9, pp. 1270 - 1280, Sep. 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....shows that we can estimate the location of moving objects and their associated velocity based on a set of initial estimates. Occluding and unlabeled regions are identified and classified in the context of a tracking algorithm. A few approaches have been adopted for solving these problems. In [1] an occlusion adaptive mesh is used for tracking moving features over several frames. In other approaches, features are extracted from a set of frames and afterwards they are tracked over the sequence. Kalman filters have been used for tracking in [2, 3, 4] Objects are segmented based on ....

Y. Altunbasak, A.M. Tekalp, "Occlusion-adaptive, content-based mesh design and forward tracking," IEEE Trans. on Image Processing, vol. 6, no. 9, pp. 1270 - 1280, Sep. 1997.

....CAMM provides an efficient representation of the image in that the number of parameters (i.e. mesh nodes) is typically much less than the number of required pixels voxels. In addition, a mesh model can also be used for motion tracking in an image sequence, by allowing the mesh to deform over time [7]. The potential benefits of using a CAMM for image reconstruction are: 1) a CAMM greatly reduces the number of unknowns, thus alleviating both the underdetermined nature of the reconstruction problem and the data storage requirement, particularly for the case of 3D reconstruction; 2) this ....

Y. Altunbasak and A. M. Tekalp, "Occlusion-adaptive, content-based mesh design and forward tracking," IEEE Trans. Image Proc., vol. 6, pp. 1270-1280, 1997.

....segmentation of the image sequence. After estimating the boundaries of the moving objects and their corresponding optical flow in 2 D we want to see how they change over the next frames [1] Tracking algorithms are usually employed for estimating the moving object position in the following frames [4, 5]. Certain regions do not have a clear match in the previous frame [1, 4] After finding the undetermined regions, their component pixels are classified by means of the already trained MRBF network. Consequently, the moving objects are matched from one frame to the next one. After estimating the ....

.... the moving objects and their corresponding optical flow in 2 D we want to see how they change over the next frames [1] Tracking algorithms are usually employed for estimating the moving object position in the following frames [4, 5] Certain regions do not have a clear match in the previous frame [1, 4]. After finding the undetermined regions, their component pixels are classified by means of the already trained MRBF network. Consequently, the moving objects are matched from one frame to the next one. After estimating the 3 D description of the moving sequence, we can derive a prediction model. ....

Y. Altunbasak, A. M. Tekalp, "Occlusionadaptive, content-based mesh design and forward tracking," IEEE Trans. on Image Processing, vol. 6, no. 9, pp. 1270 - 1280, Sep. 1997.

....and more powerful model provides a good approximation to 3 D scene structure (implicitly encoded in affine parameters of each patch) and local motion (deformations) at the expense of mesh design and tracking. The mesh model has found applications in video coding, manipulation, and indexing [12] [14]. C. Perspective Model The perspective model captures the perspective projection of the rigid motion (rotation and translation) of 3 D planar surface using eight parameters (up to a common scale) The perspective transform (homography) is given by (4) At least four corresponding point pairs are ....

Y. Altunbasak and A. M. Tekalp, "Occlusion-adaptive, content-based mesh design and forward tracking," IEEE Trans. Image Processing, vol. 6, pp. 1270--1280, Sept. 1997.

.... connectivity (a set of triangular elements) and motion (each node is attributed a sequence of motion vectors describing its temporal trajectory) information [4] Other applications of 2D dynamic meshes include video object compression, tracking, and filtering, e.g. warping motion compensation [5], 6] 7] 8] 9] 10] and video indexing and retrieval, e.g. motionbased queries [11] 3D meshes are an elementary building block of the Virtual Reality September 14, 1998 DRAFT IEEE TRANSACTIONS ON CSVT, VOL. XX, NO. Y, MONTH 1998 2 Modeling Language (VRML) a standard for storing and ....

....The motion data associated with the 2D dynamic mesh representation is determined by estimating forward motion vectors for mesh node point sequentially from one frame to the next, which is generally referred to as forward tracking. Various techniques have been proposed for to achieve this [22] [5], 8] 23] 10] Although multi resolution and hierarchical motion estimation methods have been discussed widely in literature to improve efficiency and accuracy of motion estimation, hierarchical mesh based motion estimation has been addressed only for the case of uniform meshes [7] 2] Here, ....

[Article contains additional citation context not shown here]

Y. Altunbasak and A.M. Tekalp, "Occlusion-adaptive, content-based mesh design and forward tracking," IEEE Transactions on Image Processing, vol. 6, no. 9, pp. 1270--1280, Sept. 1997.

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