Pill Recognition

Projects

Technology

Panorama SLAM

Planar SLAM

Model-Based Tracking

Marker-Based Tracking

Software Libraries

More

The Handheld Augmented Reality
Project is supported by the
following institutions:

Participate to ISMAR 2012

Follow us on twitter

Pill Recognition

Project Description

In late 2008 a Master thesis project was started to create an application for online segmentation and recognition of medicals on a mobile phone. The application should help medical stuff in recognizing unknown medicals, especially when treating patients in home care.

The application is heavily based on computer vision algorithms for segmentation of pills. Features for recognition of medicals include the color, shape, and the approximate size amongst others.

Graphical User Interface

A graphical user interface allows a convenient way for controlling the application.

On startup, the special target is detected, the
individual pills are segmented and their shape
and size are estimated.

Each individual pill can be selected, and detailed
information is given.

After querying the remote database and fetching
the results, detailed descriptions about the pills
are shown, including a sample picture of a
possible match for visual verification.

Finally the program is terminated.

Project Status

The application was developed on a standard desktop computer, and is currently ported to mobile phones. The algorithms are still under development, as well as the technical documentation is currently being written.

Project Team

Design and Implementation: Andreas Hartl

Supervision: Clemens Arth, Daniel Wagner

website maintained by Tobias Langlotz
last updated on 2011-08-02

Augmented Reality Anywhere and Anytime

Pill Recognition

Project Description

Graphical User Interface

Project Status

Project Team

copyright (c) 2011 Graz University of Technology