Title :
Robust automated footage analysis for professional media applications
Author :
Mateer, J.W. ; Robinson, J.A.
Author_Institution :
York Univ., UK
Abstract :
We report a method for automated video indexing and shot characterization that meets the specific requirements of professional post-production and archivist end users. ASAP - Automated Shot Analysis Program - interprets source video material in a manner consistent with industry practice and generates logs and searchable databases of cut location and camera activity. It uses projective transform estimation methods in conjunction with temporal filtering to resolve complex subject motion. Using challenging test footage and rigorous metrics we show that ASAP is more robust than well-established colour histogram boundary detection methods and effective at parsing complex camera movement. These results indicate that our techniques are potentially valuable for professional application.
Keywords :
database indexing; motion estimation; professional aspects; video databases; video signal processing; ASAP; Automated Shot Analysis Program; archivist end users; automated video indexing; automated video shot characterization; camera activity; colour histogram boundary detection methods; complex camera movement parsing; complex subject motion; cut location; logs; metrics; professional media applications; professional post-production; projective transform estimation methods; robust automated footage analysis; searchable databases; source video material interpretation; temporal filtering;
Conference_Titel :
Visual Information Engineering, 2003. VIE 2003. International Conference on
Print_ISBN :
0-85296-757-8
DOI :
10.1049/cp:20030493