ProceedingsComputer Society Press of the IEEE, 1993 - Computer vision |
From inside the book
Results 1-3 of 66
Page 27
... labels ) to be extracted at each instant t are : • discrete label field e = { e ( s ) , s E S } . The label set is given by = { Tm , T ,, Tp ) ; where label I'm stands for moving object , label г , for static back- ground , and label г ...
... labels ) to be extracted at each instant t are : • discrete label field e = { e ( s ) , s E S } . The label set is given by = { Tm , T ,, Tp ) ; where label I'm stands for moving object , label г , for static back- ground , and label г ...
Page 28
... label static background I , is inhibited . The choice of a 0.5 value instead of 0 allows to equally fa- vor this label when the binary observation vector ōk ( s ) has zero or only one non null component . This enables to sweep out noisy ...
... label static background I , is inhibited . The choice of a 0.5 value instead of 0 allows to equally fa- vor this label when the binary observation vector ōk ( s ) has zero or only one non null component . This enables to sweep out noisy ...
Page 29
... label maps at t = 12 and 54 Figure 9 shows motion detection labels on another infrared sequence . The latter consists of a maritime horizon with two stationary boats in the background and an approaching plane close to the center of the ...
... label maps at t = 12 and 54 Figure 9 shows motion detection labels on another infrared sequence . The latter consists of a maritime horizon with two stationary boats in the background and an approaching plane close to the center of the ...
Contents
Robust Computation of Optical Flow in a MultiScale Differential | 12 |
FigureGround | 32 |
Using Causal Semantics to Direct Focus of Attention | 49 |
Copyright | |
43 other sections not shown
Other editions - View all
Common terms and phrases
albedo algorithm analysis angle applied approach approximation B-spline boundary camera Cepstrum color constancy component Computer Vision constant constraints contour coordinates corresponding curvature curve defined deformable depth map derivatives described detection detector diagonal matrix diffuse disparity distance edge edge detection element equation error estimate Figure filter frame function fusion Gaussian geometric given global gradient IEEE illumination input invariant iterative Kalman filter label light source linear luminous intensity Machine Vision matching mathematical morphology measure method motion motion detection node noise object recognition obtained occlusion optical flow orientation parameters patch Pattern perception performance pixels planar plane position problem Proc projection pyramid reflectance region representation robust rotation scale scene scheme segments sensor sequence shape from shading shown shows smooth solution space spatial specular highlights stereo structure surface symmetry technique threshold tion tracking Trans transformation values vector velocity vergence visual