Segmentation-based multi-class semantic object detection

In this paper we study the problem of the detection of semantic objects from known categories in images. Unlike existing techniques which operate at the pixel or at a patch level for recognition, we propose to rely on the categorization of image segments. Recent work has highlighted that image segme...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Multimedia tools and applications Jg. 60; H. 2; S. 305 - 326
Hauptverfasser: Vieux, Remi, Benois-Pineau, Jenny, Domenger, Jean-Philippe, Braquelaire, Achille
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Boston Springer US 01.09.2012
Springer Verlag
Schlagworte:
Categories
Classification
Computer Communication Networks
Computer Science
Data Structures and Information Theory
Multimedia
Multimedia Information Systems
Recognition
Segmentation
Segments
Semantics
Special Purpose and Application-Based Systems
Late fusion
SVM
Segmentation
Relaxation labelling
Object detection
segmentation
bag-of-regions
CBIR
ISSN:1380-7501, 1573-7721
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:In this paper we study the problem of the detection of semantic objects from known categories in images. Unlike existing techniques which operate at the pixel or at a patch level for recognition, we propose to rely on the categorization of image segments. Recent work has highlighted that image segments provide a sound support for visual object class recognition. In this work, we use image segments as primitives to extract robust features and train detection models for a predefined set of categories. Several segmentation algorithms are benchmarked and their performances for segment recognition are compared. We then propose two methods for enhancing the segments classification, one based on the fusion of the classification results obtained with the different segmentations, the other one based on the optimization of the global labelling by correcting local ambiguities between neighbor segments. We use as a benchmark the Microsoft MSRC-21 image database and show that our method competes with the current state-of-the-art.
Bibliographie:ObjectType-Article-2
SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 23
ObjectType-Article-1
ObjectType-Feature-2
ISSN:1380-7501
1573-7721
DOI:10.1007/s11042-010-0611-2