Title :
An Optimal Method for Dynamic Particle Image Sequence
Author_Institution :
Coll. of Inf. & Electron. Eng., Zhejiang Gongshang Univ., Hangzhou, China
Abstract :
In this paper, a new optimized method with estimation and competition is proposed for false matching in movement estimation based on vision measurement technology. Firstly, the former vector results processed by optimized phase-correlation are used to estimate current matching. Secondly, the estimation results supply important parameters for competition matching. Furthermore, the method can be processed without any false modification and can restrain the multi-peak problem. At last, the method is realized in this paper and synthetic images are tested. The experimental results show that the proposed method is very effective.
Keywords :
computer vision; feature extraction; image matching; image sequences; motion estimation; dynamic particle image sequence; false matching; feature extraction; movement estimation; multipeak problem; optimal method; synthetic image; vision measurement technology; Data mining; Digital images; Educational institutions; Feature extraction; Image edge detection; Image matching; Image sequences; Optimization methods; Particle measurements; Particle tracking;
Conference_Titel :
Image and Signal Processing, 2009. CISP '09. 2nd International Congress on
Conference_Location :
Tianjin
Print_ISBN :
978-1-4244-4129-7
Electronic_ISBN :
978-1-4244-4131-0
DOI :
10.1109/CISP.2009.5305450