Real time image se tion using FPGA and

e present in this paper the realization of a classijcation hoard, fora real-time image segmentation. The clamijcation of each pixel is completed using a real' time extraction of attribub and a geometric classi3caiion method by stress polytope training, which ensures a high decision speed (I 00 ns per pixels) and good peq%rmances. The decision operator has been integrated in the form qf a full custom circuit, and ihe extraction of parameters is perjormed using a single high density FPGA .