Title :
A New Fast Hybrid Algorithm of Background Extraction Suitable for the DSP Platform
Author_Institution :
Coll. of Telecommun. & Inf. Eng., Nanjing Univ. of Posts & Telecommun., Nanjing, China
fDate :
March 31 2009-April 2 2009
Abstract :
The paper introduces background extractions with self-adaptive update algorithm and puts forward an improved algorithm based on histogram statistic combining with multi-frame average. It avoids the image trail-blur phenomena using pure multi-frame method on traffic jam, and has relative low computation complexity comparing with the hybrid Gauss model. It can run on the TI DM642 DSP hardware platform, the experimental image shows the fast extraction speed and gets good background image.
Keywords :
automated highways; computational complexity; digital signal processing chips; feature extraction; object detection; road traffic; road vehicles; statistical analysis; traffic engineering computing; video signal processing; TIDM642 DSP hardware platform; background extraction; computation complexity; histogram statistical analysis; image trail-blur phenomena; intelligent traffic system; multiframe average method; road traffic jam; road vehicle; self-adaptive update algorithm; video signal detection; Cities and towns; Data mining; Digital signal processing; Layout; Microwave technology; Motion detection; Road transportation; Road vehicles; Traffic control; Vehicle detection; Background Extraction; Intelligent Traffic System (ITS); Self-adaptive Update Algorithm; Video Detection;
Conference_Titel :
Computer Science and Information Engineering, 2009 WRI World Congress on
Conference_Location :
Los Angeles, CA
Print_ISBN :
978-0-7695-3507-4
DOI :
10.1109/CSIE.2009.406
