Product design patents retrieval technology has a great significant for patents protection. Nowadays the patent database is commonly composed of patent text description and hex-vision images. However, since the hex-vision images cannot contain all the information of the product design patents, 3D models is an important part of product design patents in future. But only text retrieval can be applied for the patent database now, which does not make full use of the hex-vision images and 3D models. Therefore, the database cannot satisfy the requirement of the retrieval and protection for product design patents. A novel hybrid search method of product design patents is proposed. The hybrid method makes use of the 3D models based on shape distribution function and takes the advantages of product hex-vision images based on both the shape index improved and the combines colour pair. The result of the experiment indicated that the proposed hybrid method can inquire and retrieval the product design patents more effectively and contribute to much for the use and protection of the product design.
[1]
Jitendra Malik,et al.
Shape contexts enable efficient retrieval of similar shapes
,
2001,
Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.
[2]
Tsuhan Chen,et al.
Indexing and retrieval of 3D models aided by active learning
,
2001,
MULTIMEDIA '01.
[3]
Taku Komura,et al.
Topology matching for fully automatic similarity estimation of 3D shapes
,
2001,
SIGGRAPH.
[4]
Thomas S. Huang,et al.
A novel relevance feedback technique in image retrieval
,
1999,
MULTIMEDIA '99.
[5]
Bernard Chazelle,et al.
Shape distributions
,
2002,
TOGS.
[6]
Anil K. Jain,et al.
Image retrieval using color and shape
,
1996,
Pattern Recognit..