implementing shot boundary (wipe) detection algorithm with matlab
Ältere Kommentare anzeigen
Please help me to impliment these five steps. thanks in advance
Read the desired number of frames from the video. Convert each frame to grey scale and resize to 256 × 256 pixel size.
Divide each frame into 4 × 4 pixel block and find the mean of each block, which will convert every frame into 64 × 64 pixel size. Consider this image as a statistical image (SI).
Calculate the mean of statistical image difference (MOSID) between consecutive frames by using MOSID = 1/(64 * 64) summation(i=1 to 64) summation(j=1 to 64) [SI(i, j, k) − SI(i, j, k + 1)]
where k varies from 1 to NFR-1, and NFR is the number of frames used for the analysis from the video
Find local threshold ST as a linear combination of m and s by ST = 0.125*m+0.125*s where m and s are mean and standard deviation of MOSID for all frames
If MOSID(k) is greater than ST, for continuous eight frames or above,then frame numbers should be printed.
Please help with the matlab implementation
Antworten (1)
Image Analyst
am 28 Okt. 2015
0 Stimmen
Try blockproc(). Demos attached.
Kategorien
Mehr zu Video Formats and Interfaces finden Sie in Hilfe-Center und File Exchange
Produkte
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!