Dynamic target classification in wireless sensor networks

Feature extraction and classification are two important components in pattern recognition. In this paper, we propose dynamic target classification in WSNs. The main idea of this approach is to dynamically switch the optimal combination of features and classifiers among different targets based on the previous result and the hypothesis during the classification process. Our hypothesis is that for each type of target, there exists an optimal combination of features and classifiers which can yield the best performance using least amount of features. We use two data sets to validate our hypothesis and derive the optimal combination sets by minimizing the cost function. Then we apply the proposed algorithm to a scenario of target classification to illustrate how it works in WSNs. Numerical results show that our approach can significantly reduce the computational complexity, and at the same time, achieve the better accuracy compared with the normal classification, making it an attractive cho...

• Type: Conference
• Year: 2008
• Where: ICPR
• Authors: Ying Sun, Hairong Qi
• Keywords: Computer Vision, Dynamic Target Classification, ICPR 2008, Optimal Combination, Optimal Combination Sets