EVANDRO OTTONI - Academia.edu (original) (raw)
Uploads
Papers by EVANDRO OTTONI
Image representation is an important issue for many visual computing tasks such as image classifi... more Image representation is an important issue for many visual computing tasks such as image classification, understanding, feature extraction and retrieval. This paper presents a new image descriptor for scenes classification, called CMCT (Contextual Mean Census Transform), that combines distribution of local structures with contextual information. Ex-perimental results on commonly used datasets demonstrate that the proposed method could achieve competitive performance against previous methods.
Image representation is an important issue for many visual computing tasks such as image classifi... more Image representation is an important issue for many visual computing tasks such as image classification, understanding, feature extraction and retrieval. This paper presents a new image descriptor for scenes classification, called CMCT (Contextual Mean Census Transform), that combines distribution of local structures with contextual information. Ex-perimental results on commonly used datasets demonstrate that the proposed method could achieve competitive performance against previous methods.