Skip to Main content Skip to Navigation
Theses

Quantification vectorielle algébrique avec zone morte : application à la compression d'images à bas débit et au tatouage d'images

Abstract : We propose to use Lattice Vector Quantization (LVQ), which combines the main advantages of vector approaches and a low complexity. The originality of our method consists in defining a vectorial dead zone, adapted to codebook shape and associated with LVQ. Furthermore, a theoritical rate (i.e. compression ratio) model has been developed, in order to accelerate the optimal setting of the dead zone shape parameter. Moreover, the redundancy generated by the reset to zero of numerous coefficient blocks has been used with the Stack-Run coder. Our results have been compared with the current reference algorithms, namely JPEG2000 and SPITH. In another field, Dead Zone Lattice Vector Quantization has been used very relevantly for both watermaking and compression, permitting to improve the watermaking robustness to some attacks such as an additive gaussian noise.
Complete list of metadatas

Cited literature [97 references]  Display  Hide  Download

https://hal.univ-lorraine.fr/tel-01748638
Contributor : Thèses Ul <>
Submitted on : Thursday, March 29, 2018 - 11:43:46 AM
Last modification on : Wednesday, April 11, 2018 - 1:28:09 AM
Long-term archiving on: : Friday, September 14, 2018 - 5:00:40 AM

File

SCD_T_2003_0033_VOINSON.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-01748638, version 1

Collections

Citation

Teddy Voinson. Quantification vectorielle algébrique avec zone morte : application à la compression d'images à bas débit et au tatouage d'images. Autre. Université Henri Poincaré - Nancy 1, 2003. Français. ⟨NNT : 2003NAN10033⟩. ⟨tel-01748638⟩

Share

Metrics

Record views

80

Files downloads

483