摘要
以全向视觉节点为基础,研究视觉网络中节点定位精度与效率问题。全向视觉网络中各个锚节点测量未知节点得到方位角。通过分析两个不同锚节点所测得的方位角之差对定位误差传递的影响,提出锚节点对(Anchor Node Pair,ANP)定位优先度与权值概念。定位ANP之间根据优先度参与定位竞争,利用在竞争中被选出的ANP所测方位角进行未知节点位置估计。融合算法将多个定位结果权值融合,得到目标定位结果。实验结果表明,算法在多种条件下都能避免夹角过大、过小引起的定位误差发散的问题,提高了定位精度与稳定性。
In background of omni-directional vision node, discuss accuracy and efficiency of visual network node location. Each anchor node measures unknown node azimuth. By analyzing the azimuth difference impact to location error propagation, propose location priority and weight of the Anchor Node Pair（ANP）. The anchor node pairs participate in location competition with priority. The algorithm weights the mid-results to get the final fusion location. Through simulation, the algorithm is verified effective, which avoids location inaccuracy, while the azimuth difference is too large or small and improves the location accuracy and efficiency of visual sensor network target.
出处
《计算机工程与应用》
CSCD
2014年第10期147-151,共5页
Computer Engineering and Applications
基金
国家自然科学基金(No.60973095)
关键词
全向视觉
锚节点对
方位角
优先度
融合定位
omni-directional vision
anchor node pair
azimuth
priority
fusion location
