We propose to incorporate a priori geometric constraints in a 3?D stereo reconstruction scheme to cope with the many cases where image information alone is not sufficient to accur...
– This paper addresses the exploitation of massive numbers of image-derived change detections. We use the term “change analysis” to emphasize the intelligence value obtained ...
Computing high quality 3D models from multi-view stereo reconstruction is an active topic as can be seen in a recent review [15]. Most approaches make the strong assumption that th...
We propose the use of latent space models applied to local invariant features for object classification. We investigate whether using latent space models enables to learn patterns...
Florent Monay, Pedro Quelhas, Daniel Gatica-Perez,...
This paper describes an approach to the use of gradient descent search in genetic programming (GP) for object classification problems. In this approach, pixel statistics are used ...