There are many scenarios where the liquid is occluded by other items (e.g. books in a packet), in which existing RF-based liquid identification methods are generally not suitable. Moreover, status methods are not applicable when the height of the liquid to be tested changes. This paper proposes PackquID, an RF-based in-packet liquid identification system, which can identify liquid without prior knowledge. In dealing with the obstruction of other items and the unknown container, we utilize a dual-antenna model and craft a relative frequency response factor, exploring the diversity of the permittivity in the frequency domain. In tackling the variable liquid height, we extend our model to 3D scope by analyzing the electric field distribution and solving the height effect via spatial-differential model. With 500 pages of printer paper obscured, PackquID can identify 9 common liquids, including Coca-Cola and Pepsi, with an accuracy of over 86% for 4 different packets (canvas bag, paper bag, backpack, and box) and 4 different containers. Nevertheless, PackquID can still identify liquids with an accuracy rate of over 87%, even when the liquid height changes from 4 cm to 12 cm. https://dl.acm.org/doi/10.1145/3569469

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/papers/ubicomp/128324/2023

AdRecommended

Learn AI Coding at CodeNow

At a Glance

Paper Snapshot

fact_check
dataset
Source
UbiComp
calendar_month
Year
2023
emoji_events
Award
No award tagged
group
Authors
4 authors
sell
Subtopics
Context-Aware Computing, Ubiquitous Computing
work
Professions
—
article
Content Status
Abstract only
hub
Related Papers
10 related papers