Event Detection for Content-based Multimedia Retrieval
碩士 === 輔仁大學 === 電子工程學系 === 91 === Content-based multimedia retrieval is a technology that automatically transforms multimedia content into indices for the purpose of inquiry. Early studies often used low-level image features as indices. This paper studies scene segmentation in film indexi...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2003
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/32805268936737240673 |
| Summary: | 碩士 === 輔仁大學 === 電子工程學系 === 91 === Content-based multimedia retrieval is a technology that automatically transforms multimedia content into indices for the purpose of inquiry. Early studies often used low-level image features as indices. This paper studies scene segmentation in film indexing system and organization of content after indexing for fast query. The approach: 1) shot boundaries detection; 2) key frame extraction; 3) features extraction; 4) event inference. Based on the 4 stages, we can automate scene analysis and segmentation of a film. Experimental results have shown positive outcomes.
|
|---|