Parameter Estimation for Non-convex Target Object Using Networked Binary Sensors

Hiroshi SAITO  Sadaharu TANAKA  Shigeo SHIODA  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E94-D   No.4   pp.772-785
Publication Date: 2011/04/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.E94.D.772
Print ISSN: 0916-8532
Type of Manuscript: PAPER
Category: Fundamentals of Information Systems
Keyword: 
parameter estimation,  shape estimation,  sensor network,  ubiquitous network,  integral geometry,  

Full Text: PDF>>
Buy this Article




Summary: 
We describe a parameter estimation method for a target object in an area that sensors monitor. The parameters to be estimated are the perimeter length, size, and parameter determined by the interior angles of the target object. The estimation method does not use sensor location information, only the binary information on whether each sensor detects the target object. First, the sensing area of each sensor is assumed to be line-segment-shaped, which is a model of an infrared distance measurement sensor. Second, based on the analytical results of assuming line-segment-shaped sensing areas, we developed a unified equation that works with general sensing areas and general target-object shapes to estimate the parameters of the target objects. Numerical examples using computer simulation show that our method yields accurate results.