Author: Mohammadhassan REZAEI, Erkan GUNPİNAR
Publishing Date: 2018
E-ISSN: 2147-9364
Volume: 6 Issue: 1
ABSTRACT:
Due to the large size of shape databases, importance of effective and robust method in shape retrieval has been increased. Researchers mainly focus on finding descriptors which is suitable for rigid models. Retrieval of non-rigid models is a still challenging field which needs to be studied more. For non-rigid models, descriptors that are designed should be insensitive to different poses. For non-rigid model retrieval, we propose a new method which first divides a model into clusters using geodesic distance metric and then computes the descriptor using these clusters. Mesh segmentation is performed using a skeleton-based K-means clustering method. Each cluster is represented by an area based descriptor which is invariant to scale and orientation. Finally, similar objects for the input model are retrieved. Articulated objects from human to animals are used for this study’s experiments for the validation of the proposed retrieval algorithm.
Keywords: Shape Retrieval; K-means Clustering; Mesh Skeleton; Geodesic distance