作者: Peng Huang , Adrian Hilton
DOI: 10.1109/ICCVW.2009.5457434
关键词: Computer animation 、 Curve fitting 、 Pattern recognition 、 Mathematics 、 Image retrieval 、 Character animation 、 Animation 、 Object (computer science) 、 Matching (statistics) 、 Histogram 、 Computer vision 、 Artificial intelligence
摘要: Most 3D object retrieval and matching methods only consider geometric similarity. This paper introduces a novel descriptor, Shape-Colour Histograms, to match objects with similar shape appearance. is motivated by the requirement concatenate captured video sequences for animation production. A quantitative evaluation based on Receiver-Operator Characteristic (ROC) curve presented compare performance of conventional descriptors new shape-colour temporal filtering in task sequences. including Shape Histogram, LightField are considered. Evaluation shows that filtered Histograms outperform using only. Finally, applied publically available database people identify optimal transitions synthesize character animation. Results demonstrate accurate surface shape, appearance motion at transition points.