Fast segmentation, tracking, and analysis of deformable objects

Abstract

To cite this version:Chahab Nastar, Nicholas Ayache. Fast segmentation, tracking and analysis of deformable objects.

Modal Space: A Physics-Based Model for Sequential Estimation of Time-Varying Shape from Monocular Video

“…The FEM was proposed to accurately represent specific materials (Tsap et al 2000;Young and Axel 1993) with known material properties. To tackle the high dimensionality of these physics-based models, a low-rank representation was proposed by applying modal analysis over a known structure discretized in 3D finite elements (Nastar and Ayache 1993;Pentland and Horowitz 1991). This method was then applied to image segmentation (Pentland and Sclaroff 1991), medical imaging (Nastar and Ayache 1993) and deformable 2D motion tracking (Tao and Huang 1998;Sclaroff and Pentland 1995).…”

“…To tackle the high dimensionality of these physics-based models, a low-rank representation was proposed by applying modal analysis over a known structure discretized in 3D finite elements (Nastar and Ayache 1993;Pentland and Horowitz 1991). This method was then applied to image segmentation (Pentland and Sclaroff 1991), medical imaging (Nastar and Ayache 1993) and deformable 2D motion tracking (Tao and Huang 1998;Sclaroff and Pentland 1995). Pentland and Horowitz 1991 integrated the incremental updates of the basis coefficients over time using an EKF.…”

“…Modal analysis is standard in structural engineering and has also been applied in computer vision to the spring mesh model in medical imaging (Pentland and Sclaroff 1991). In (Nastar and Ayache 1993;Pentland and Horowitz 1991) was used for motion analysis and to track and recover heart motion, and in (Sclaroff and Pentland 1995;Tao and Huang 1998) for non-rigid 2D tracking. Modal analysis was used to decouple the equilibrium equations by obtaining a closedform solution of Eq.…”

Consistency of Optimized Facial Features through the Ages

“…There are many techniques are applied to extract the features of the face during that time [1][2][3][4][5][6][7][8][9][10][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42]. However, a little number of papers is discussed or discovered how the features of a face are related to each other for a specific person.…”

Simultaneous Surface Approximation and Segmentation of Complex Objects

“…Several variants of snakes exist which are based on Fourier descriptors [38,39], B-splines [3,40,41], finite elements [42], balloons [43], and discrete representation [44,45], and the literature on 3-D deformable models is also rich [11,[19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37]48]. Good results have been obtained, but these algorithms handle mostly topologically simple objects and might fail under the following two conditions: -First, there may be more than one underlying object, and these objects might be close to one another.…”