Shape Distribution-Based 3D Shape Retrieval Methods: Review and Evaluation

Three-dimensional (3D) models are basis of modern digital manufacturing systems. It has attracted much attention in research and industrial applications to build 3D product models efficiently and cost-effectively. One of promising developments is product models reusing through sharing. The efficient retrieval of 3D product shapes can speed up 3D modeling for product design and manufacturing but it remains a technical challenge to search for similar 3D models or parts in a model set. Traditional text-based search engines fail in 3D model searching as the lack of detailed descriptions. This paper reviews current research and development in shape distribution-based 3D model retrieval methods. The methods are evaluated by the comparison of their efficiency and accuracy using examples. The further research is proposed based on solutions found.