METHOD FOR CODING AND DECODING A SIGNAL DEFINED ON A VARIETY OF DIMENSION N, DEVICE FOR CODING AND DECODING SUCH A SIGNAL AND CORRESPONDING COMPUTER PROGRAMS

摘要

The invention relates to the coding of a video signal (f) representative of the geometry of an object defined on a space of dimension n (n≥1), comprising the steps consisting in: - representing (C1) said object with the aid of a series of meshes M₀, Μ₁,...Μ_i-1, M_i, M_i+1,..., Ms (0≤i≤S) formed respectively of a plurality of simplexes of dimension n, a current mesh considered M, constituting a subdivision of the immediately preceding mesh Μ_i-1, - projecting (C2) said signal in a basis of wavelets into a plurality of signal portions f₀, f₁,..., f_i-1, f_i, f_i+1,..., f_s, a portion f_i, of said signal being defined, for said current mesh considered M_i, as the affine function over each simplex of said mesh M_i, - applying (C3) to said portion f_i of said signal a wavelet transformation, delivering a wavelet coefficient D_i-1, - iterating the application steps for each of the signal portions, delivering a set of wavelet coefficients D₀, D₁; D_s-1, - coding (C5) the wavelet coefficients delivered. In the case where i is even, the subdivision of the mesh M_i, to obtain the mesh M_i+1, is performed according to a first predetermined subdivision rule, and in the case where i is odd, the subdivision of the mesh M_i, to obtain the mesh M_i+1, is performed according to a second predetermined subdivision rule.