The Moving Target Detection under Complicated Background Base on BP Neural Networks

Abstract:

This paper aims at utilizing BP neural network and improved BP neural network in the extraction of moving targets from the static background and the motor background. Across introduces of L_M Optimization method to improve the BP neural network, combination the difference between the consecutive frames or difference from the background for the extraction of moving targets. The paper uses one types of image segmentation improved BP neural network for extraction of moving targets from the static background. We will obtain the same moved targets with the extraction of moving targets from the static background and the motor background. After experimental verification, this method on different objects in the context of the same detection ability is better.

• Periodical: Advanced Materials Research (Volumes 706-708)
• Pages: 1072-1076
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.706-708.1072
• Online since: June 2013
• Authors: Xiang Dong Feng
• Keywords: BP Neural Network (BPNN), Image Segmentation, L_M Method, Moving Target Detection

References

[1] Haritaoglu I, Larry S.Davis and D.Harwood. A real time system for detecting and tracking people[C].Japan: International conference on face and gesture recogntion, 1998.

DOI: 10.1109/afgr.1998.670952

[2] Ridder C, Munket O, Kirchner H. Adaptive background estimation and Foreground detection using kalman-filtering [J].Proceeding of International Conference on Recent Advances in Macaronis, 1995, 1:193-199.

[3] Cucchiarar, Granac, Piccardim. Detecting moving objects, ghosts, and Shadows in video streams [J].IEEE Transactions on Pattern Analysis and machine Intelligence.2003, 25(10): 1337 -1342.

DOI: 10.1109/tpami.2003.1233909

[4] K.V. Mardia T.J. Hainsworth. A spatial threshold method for image segmentation [J].IEEE Transactions on Pattern Analysis and MachineIntelligence.1998, 10: 219-237.