Title :
Integration of Multi-feature for Moving Target Detection Algorithm
Author :
Yu, Jin ; Qian, Feng
Author_Institution :
Key Lab. of Adv. Control & Optimization for Chem. Processes, East China Univ. of Sci. & Technol., Shanghai, China
Abstract :
Moving object segmentation and marking object area is one of the key technologies of intelligent surveillance. In this paper, combining the background modeling in pixel level, the proposed algorithm presented gradient-based spatial characteristics in order to improve detection accuracy. For the trade-off problem of background modeling, established the velocity characteristics for each property of moving targets, to make online updates of the GMM more adaptable. The results indicate that this technique can improve the efficiency and stability of detection.
Keywords :
object detection; surveillance; gradient based spatial characteristic; intelligent surveillance; moving target detection algorithm; object segmentation; Adaptation model; Computational modeling; Computer vision; Hidden Markov models; Image segmentation; Object detection; Pixel; Background subtraction; Gradient feature; Intelligent surveillance; Shape feature; Velocity feature;
Conference_Titel :
Electrical and Control Engineering (ICECE), 2010 International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-6880-5
DOI :
10.1109/iCECE.2010.1282
