Key frames extraction for video
Date of Issue2016-05-17
School of Electrical and Electronic Engineering
Key frames are important information in video management that allow us to extract notable information from the video, and reduce the transmission data of communication network. A typical video normally consists of 20 to 30 frames per second. The amount of data in video is significantly reduced by using key frames extraction. In this project, a MATLAB program for video key frames extraction is developed. Key frames extraction based on color histogram is discussed in this report. Calculations on the color histogram difference between consecutive frames are conducted. For color histogram difference method, when a difference exceeds the threshold value, a key frame will be extracted. In addition, 3D histogram representations of the key frames will be produced for further analysis and processing. The experimental results will be presented and discussed.
Final Year Project (FYP)
Nanyang Technological University