A novel machine learning approach in Image Pattern Recognition under invariance constraints - IMEP-LAHC - Institut de Microélectronique, Electromagnétisme et Photonique - Laboratoire d'Hyperfréquences et Caractérisation Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

A novel machine learning approach in Image Pattern Recognition under invariance constraints

Résumé

Two of the main challenges of image recognition in radar, acoustic or Tray imaging regard the viewpoint variation of the pattern and the feature extraction techniques that must retrieve the most discriminative information about different classes. In this paper, we focus on feature extraction and image classification techniques by using a Rotation Invariant Wavelet Packet Decomposition and a novel entropybased feature extraction technique to characterize an image. The entropy-based characterization described in the paper offers an extended analysis compared to usual approaches such as the energy of the wavelet sub bands. The computed features will be further used to train a Graph Neural Network adapted to a quad-tree decomposition which has the powerful advantage of considering the structural information of the rotationinvariant decomposition. We successfully classified the images with an accuracy of 99.3%. The results are compared to other classic feature extraction techniques such as k-NN, SVM and WPD, proving the increased capability of our method.
Fichier principal
Vignette du fichier
A novel machine learning approach in Image Pattern Recogniton under invariance constraints-1.pdf (676.26 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03980830 , version 1 (09-02-2023)

Identifiants

Citer

Dragos Nastasiu, Angela Digulescu, Cornel Ioana, Maxime Bernier, Frederic Garet, et al.. A novel machine learning approach in Image Pattern Recognition under invariance constraints. EUSIPCO 2022 - 30th European Signal Processing Conference, Aug 2022, Belgrade, Serbia. pp.558-562, ⟨10.23919/EUSIPCO55093.2022.9909907⟩. ⟨hal-03980830⟩
12 Consultations
31 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More