Skip to Main Content
Frequently Asked Questions
Submit an ETD
Global Search Box
Need Help?
Keyword Search
Participating Institutions
Advanced Search
School Logo
Files
File List
Sai Luo.pdf (2.24 MB)
ETD Abstract Container
Abstract Header
Semantic Movie Scene Segmentation Using Bag-of-Words Representation
Author Info
luo, sai
Permalink:
http://rave.ohiolink.edu/etdc/view?acc_num=osu1500375283397255
Abstract Details
Year and Degree
2017, Master of Science, Ohio State University, Civil Engineering.
Abstract
Video segmentation is an important procedure in video indexing and archiving. Shot and scene are two levels of video segment. In this thesis, a shot detection method and a scene segmentation method are introduced. The shot detection method is based on both camera motion and histogram difference in the frames. Motions in the frames within one shot have certain continuity, so the frames which break the pattern are the shot transition frames. Motion based method is sensitive to object motion in frames, which would affect the result, thus histogram comparison is introduced. Frames in one shot are very similar in terms of visual content, the visual content can form the histogram. The drawback of this method is that in scenes with massive camera movement, the visual content may vary much. By combining these two methods the shots can be extracted with higher accuracy. Scene segmentation is a higher level in video summarization. In this process, BOW model and color histogram method are used. BOW model is used to construct feature descriptors to demonstrate background information. The model is constructed by generating clusters from the feature vectors of the training images. Since scene is a complex level of video segment, we need to use multiple criteria to determine it, thus color histogram is introduced. Color information of the frames in one scene shares similarity, the frames which break the pattern are the frames with scene transition, thus forming the second level scenes.
Committee
Alper Yilmaz (Advisor)
Wang Lei (Committee Member)
Qin Rongjun (Committee Member)
Pages
93 p.
Subject Headings
Civil Engineering
Keywords
Shot detection, Scene segmentation, Bag-of-Words
Recommended Citations
Refworks
EndNote
RIS
Mendeley
Citations
luo, S. (2017).
Semantic Movie Scene Segmentation Using Bag-of-Words Representation
[Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1500375283397255
APA Style (7th edition)
luo, sai.
Semantic Movie Scene Segmentation Using Bag-of-Words Representation.
2017. Ohio State University, Master's thesis.
OhioLINK Electronic Theses and Dissertations Center
, http://rave.ohiolink.edu/etdc/view?acc_num=osu1500375283397255.
MLA Style (8th edition)
luo, sai. "Semantic Movie Scene Segmentation Using Bag-of-Words Representation." Master's thesis, Ohio State University, 2017. http://rave.ohiolink.edu/etdc/view?acc_num=osu1500375283397255
Chicago Manual of Style (17th edition)
Abstract Footer
Document number:
osu1500375283397255
Download Count:
963
Copyright Info
© 2017, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.