Title :
Representation Method of the Moving Object Trajectories by Interpolation with Dynamic Sampling
Author :
Jin-Hwal Lee ; Kee-Sung Lee ; Geun-Sik Jo
Author_Institution :
Saltlux Inc., Seoul, South Korea
Abstract :
In this study, we experiment the Cubic Spline Interpolation annotation method in order to reduce annotation size for stable or uncertainty cause by the object movement. In the experiment result, we found that Cubic Spline Interpolation method is able to save up to 80 percent of annotation size with 15 frames annotation and below. Moreover, it produces error rate at 31 pixels and averagely less than 4 pixels. Our proposed method to employ dynamic sampling by dynamically change the sampling rate according to object speed shows significant decreases in error rates from 69 percent to 65 percent.
Keywords :
image motion analysis; image representation; image sampling; interpolation; splines (mathematics); annotation size reduction; cubic spline interpolation annotation method; dynamic sampling rate; moving object trajectory representation method; object movement; object speed; Error analysis; Interpolation; Semantics; Splines (mathematics); Trajectory; Transform coding; Videos;
Conference_Titel :
Information Science and Applications (ICISA), 2013 International Conference on
Conference_Location :
Suwon
Print_ISBN :
978-1-4799-0602-4
DOI :
10.1109/ICISA.2013.6579375