Multi-scale Ternary and Septenary Patterns for Texture classification - Université Gustave Eiffel Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Multi-scale Ternary and Septenary Patterns for Texture classification

Résumé

This paper proposes, inspired by local binary pat- tern (LBP) and its variants, a novel local texture operator for texture modelisation and classification, referred to as Multi-scale Ternary and Septenary Pattern (MTSP). MTSP is a histogram- based feature representation designed based on two single-scale STP and SSP encoders (single-scale ternary and septenary pat- terns, respectively). STP and SSP are built using a new set theory- based pattern encoding technique that combines the concpet of both LTP’s and LQP’s texture descriptors. The main idea behind STP and SSP is to calculate several virtual pixels based on different local and global image statistics and to progressively encode both local and non-local pixel interactions by analyzing the differential excitation and direction information based on the relationships between pixels sampled in different locations. Then, the obtained histograms of SSP and TSP methods are concatenated to form the final MTSP feature vector. Experiments have shown that MTSP has better performance stability across nine texture datasets than many recent state-of-the-art texture approaches.
Fichier principal
Vignette du fichier
conference_101719.pdf (1.28 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04529576 , version 1 (02-04-2024)

Identifiants

  • HAL Id : hal-04529576 , version 1

Citer

Rachdi El Mokhtar, Youssef El Merabet,, I. El Khadiri, Y Rhazi, C Meurie. Multi-scale Ternary and Septenary Patterns for Texture classification. The 1st International conference on Artificial Intelligence: Advanced Topics and Social Issues - AIATSI, Nov 2022, Marrakesh, Morocco. ⟨hal-04529576⟩
3 Consultations
3 Téléchargements

Partager

Gmail Facebook X LinkedIn More