Manualshelf

User Manual

ManualsBrandsSeeed Studio ManualsDevelopment Boards & KitsMGC3130 3D Gesture and Tracking Shield for Raspberry Pi
12345678910
MGC3030/3130
DS40001667E-page 10 Advance Information 2012-2017 Microchip Technology Inc.
2.0 FEATURE DESCRIPTION
2.1 Gesture Definition
A hand gesture is the movement of the hand to express
an idea or meaning. The GestIC
®
technology
accurately allows sensing of a user’s free space hand
motion for contact free position tracking, as well as 3D
gesture recognition based on classified movement
patterns.
2.2 GestIC Library
MGC3X30 is being provided with a GestIC Library
loader which is stored on the chip’s Flash memory.
Using this loader, a GestIC Library can be flashed on
the MGC3X30 via I
2
C™ with (e.g., Aurea GUI) (see
Section 7.1 “Aurea Software Package”) or an
embedded host controller. The GestIC Library
includes:
• Colibri Suite: Digital Signal Processing (DSP)
algorithms and feature implementations.
• System Control: MGC3X30 hardware control
features such as Analog Front End (AFE) access,
interface control and parameters storage.
• Library Loader: GestIC Library update through the
application host’s interface.
2.2.1 COLIBRI SUITE
The Colibri Suite combines data acquisition, digital
signal processing and interpretation.
The Colibri Suite functional features are illustrated in
Figure 2-1 and described in the following sections.
FIGURE 2-1: COLIBRI SUITE CORE
ELEMENTS
2.2.1.1 Position Tracking (MGC3130)
The Colibri Suite’s Position Tracking feature provides
three-dimensional hand position over time and area.
The absolute position data is provided according to the
defined origin of the Cartesian coordinate system (x, y,
z). Position Tracking data is continuously acquired in
parallel to Gesture Recognition. With a position rate of
up to 200 positions/sec., a maximum spatial resolution
of 150 dpi is achieved.
2.2.1.2 Gesture Recognition (MGC3X30)
The Colibri Suite’s gesture recognition model detects
and classifies hand movement patterns performed
inside the sensing area.
Using advanced stochastic classification based on
Hidden Markov Model (HMM), industry best gesture
recognition rate is being achieved.
The Colibri Suite includes a set of predefined hand
gestures which contains flick, circular and symbol
gestures as the ones outlined below:
• Flick gestures
FIGURE 2-2: FLICK GESTURES
A flick gesture is a unidirectional gesture in a quick
flicking motion. An example may be a hand movement
from West to East within the sensing area, from South
to North, etc.
• Circular gestures
FIGURE 2-3: CIRCLE GESTURES
A circular gesture is a round-shaped hand movement
defined by direction (clockwise/counterclockwise)
without any specific start position of the user’s hand.
Two types of circular gestures are distinguished by
GestIC technology:
1. Discrete Circles
Discrete Circles are recognized after performing a
hand movement inside the sensing area. The
recognition result (direction: clockwise/
counterclockwise) is provided after the hand movement
stops or the hand exits the detection area. The Discrete
Circles are typically used as dedicated application
control commands.