To exploit the road network in raster maps, the first step is to extract the pixels that constitute the roads and then vectorize the road pixels. Identifying colors that represent...
Using Comaniciu et al.’s approach as a basis, [9], this paper presents a real-time tracking technique in which a multiple target model is used. The use of a multiple model shall ...
Visual learning is expected to be a continuous and robust process, which treats input images and pixels selectively. In this paper we present a method for subspace learning, which...
Automatic unloading of piled boxes of unknown dimensions is undoubtedly of great importance to the industry. In this contribution a system addressing this problem is described: a l...
This paper presents a new algorithm based on boosting for interactive object retrieval in images. Recent works propose ”online boosting” algorithms where weak classifier sets...
Alexis Lechervy, Philippe Henri Gosselin, Frederic...