Combining Co-Training and Co-Testing for Interactive Image Segmentation

Author

Wang, Liantao ; Hu, Xuelei ; Lu, Jianfeng

Author_Institution

Sch. of Comput. Sci. & Technol., Nanjing Univ. of Sci. & Technol., Nanjing, China

fYear

2012

Firstpage

596

Lastpage

599

Abstract

This paper presents an elastic framework with human interactions for image segmentation. Under this framework, we extract color and texture features for each pixel and consider them as two independent views. Based on pixel classification, the image segmentation can be accomplished through combining multi-view semi-supervised and active learning with human interactions. Experiments on various color images show the feasibility and effectiveness of our framework.

Keywords

feature extraction; image classification; image colour analysis; image segmentation; image texture; interactive systems; learning (artificial intelligence); active learning; co-testing; co-training; color feature extraction; color images; elastic framework; human interactions; interactive image segmentation; multiview semisupervised learning; pixel classification; texture feature extraction; Classification algorithms; Feature extraction; Humans; Image color analysis; Image segmentation; Support vector machines; Training; co-testing; co-training; interactive image segmentation; multi-view learning;

fLanguage

English

Publisher

ieee

Conference_Titel

Industrial Control and Electronics Engineering (ICICEE), 2012 International Conference on

Conference_Location

Xi´an

Print_ISBN

978-1-4673-1450-3

Type

conf

DOI

10.1109/ICICEE.2012.162

Filename

6322451