Abstract
Recent work has shown that classification can benefit saliently from the proper use of contextual cues. In this paper, we present a novel framework by using contextual information to improve object detection. After having detected objects and clustered image regions respectively, we make the statistic spatial relationships between candidate objects and sub-regions of the image, which are composed into a vector of features (termed contextual words). Then, the vectors of all candidates are used to train the probabilistic Latent Semantic Analysis (pLSA) model. Finally, SVM classifier is employed in object classification. Furthermore, a new direct graphical model by extending pLSA is proposed to introduce more useful information from previous detection process. The experimental results based on a remote-sensing image dataset show that our performances are quite promising.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
