원문정보
초록
영어
This paper proposes a useful algorithm to count the metrological parameter center in wireless sensor network. By setting assigned threshold in Mean shift iteration, this algorithm lessens the computation load and can control the center points to be extracted. Experiments verify the efficient center finding algorithm and the control factors in center finding. This algorithm can be extended to more wide range for application. Because the mean shift is robust in convergence, the WSNs is utilized in many metrological area, the algorithm can be used widely in these domains. For the stability of image data, we use color space transform to decrease the illumination influence.
목차
1. Introduction
2. Feature Extraction using Mean Shift
3. Experiment and Analysis
4. Conclusions
References