A floor plan creation tool utilizing a smartphone with an ultrasonic sensor gadget
A floor plan creation tool utilizing a smartphone with an ultrasonic sensor gadget
复制标题
利用带有超声波传感器小工具的智能手机的平面图创建工具
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
K. Yasumoto
中科院分区：
文献类型：
--
作者：
Yukitoshi Kashimoto;Yutaka Arakawa;K. Yasumoto
这些天来，在本地室内导航的各种应用程序最近引起了极大的关注的主张是挑衅。为了改善和优化室内导航应用，我们必须获得目标场所（如建筑物和家庭环境）的精确平面图。然而，认为所有的建筑物都为我们提供了室内导航的准确平面图是毫无根据的。在本文中，我们提出了一个平面图生成器工具，利用智能手机集成超声波传感器小工具。在测量中，用户沿着所有房间的墙壁完成沿着圈。然后，系统估计房间的准确形状和大小。它利用智能手机中的惯性传感器来跟踪用户的行走路径。此外，小工具中的超声波传感器测量路径和墙壁之间的距离。要实现优化的性能，有两个主要挑战。第一个挑战是，相邻的对象，如书架，影响空间布局估计的准确性。为了科普这个问题，我们使用混合高斯滤波器。第二个挑战是狭窄的房间，如走廊，导致精度低。为了缓解这一挑战，我们在相反的方向上实现了两个超声波传感器，并直接测量墙壁之间的距离。实验结果表明，形状和尺寸估计精度有所提高。
These days, the proposition that various applications in local indoor navigation have attracted significant attention recently is the provocative one. In order to improve and optimize the indoor navigation applications, we have to obtain the accurate floor plan of target place such as building and home environments. However, the belief that all the buildings provide us with accurate floor plan for indoor navigation is groundless. In this paper, we present a floor plan generator tool utilizing a smartphone integrated with an ultrasonic sensor gadget. In the measurement, the user completes a lap along the walls of all rooms. Then, the system estimates the accurate shape and size of the rooms. It leverages the inertial sensors, implemented in the smartphone, to track the user in the walking path. Moreover, the ultrasonic sensor in the gadget measures the distance between the path and walls. There are two main challenges to achieve optimized performance. The first challenge is that adjacent objects to, such as bookshelves, affect the accuracy in spatial layout estimation. To cope with this problem, we use a mixed Gaussian filter. The second challenge is that the narrow room, such as corridors, leads to the low accuracy. To alleviate this challenge, we implement two ultrasonic sensors in the reverse direction, and measure the distance between walls directly. The results from experimentations report improvement in shape and size estimation accuracy.