Note: Descriptions are shown in the official language in which they were submitted.
CA 02432751 2003-06-20
BACKGROUND OF THE INVENTION
When an alarm system is activated, there is usually a specific sound that is
produced to warn acoustically the people about the alarming event. This is the
case with a
fire alarm and a carbon monoxide detector for example. Typically there is no
action other
than the sound that is generated, and only in a few cases a signal is also
triggered to
activate a monitoring device. This is a major disadvantage because it
significantly delays
the action taken in response to the alarming event. There are cases when the
triggering
sound events are not produced by other alarm systems {electronically
generated), but are
themselves a result of an action that someone would like to be notified on,
(e.g. the sound
generated by a breaking glass window, a barking dog, a crying child or a sick
person
seeking help). There are examples in the prior art of sy stems detecting such
types of
sounds (CA 2 117 053, CA 2 270 8I4, US 5 977 871, US 5 315285), but the object
of
this invention is a method and an apparatus that will allow the user tc~
specify the
triggering alarm sound through a training sequence. After being trained on a
set of
alarming sounds, an analysis of the incoming sound will be performed and
through a
digital signal processing algorithm its digital signature will be compared
with the ones
that have already been calculated during the training sequence initiated by
the user. When
a matching is found between two digital signatures, a notification signal {a
phone call, an
e-mail message, a flashing light or any other notification signal) will be
triggered. The
examples found in the prior art fail to address the training aspect of sound
detection, and
the ability to handle all the alarming events that happen in a unified way,
that being by
processing the sounds associated with those alarms.
CA 02432751 2003-06-20
SUM1VIARY ~F THE INVENTION
It is an object of the present invention to disclose an integrated alarm
system that
is triggered by the sound associated with an alarming event. An alarming event
is
defined in this context as being the sound generated by an existing alarm
system (fire
detecting alarm, carbon monoxide detector alarm, burglary detection alarm) or
the
sound generated by other means (a dog barking, a child crying or a disable
person
seeking help).
It is another object of the present invention to disclose an integrated alarm
monitoring
system that allows the user to train it for the purpose of recagnizing
specific sounds.
There is no requirement of knowing from before what alarm event will be needed
to
be detected. It is the user's privilege to choose what the alarming event will
be, and to
specify what sound is considered to be appropriate to generate a notification
to the
user.
It is another object of the present invention to disclose an algorithm that
calculates a
digital signature for the incoming sound and compares it with all the other
digital
signatures previously stored in the memory and to create a notification signal
over a
phone line, an e-mail message or by any other mean, when a matching is
detected.
This is achieved by employing a transducer to convert the energy of the sound
into an
electrical signal, a sampling mechanism with a sampling period smaller that
the
period of the event that is intended to be detected, a digital signal
processing
algorithm that processes the samples of the current electrical stimulus
representing
the result of the sampling process and calculates the probability of the
stimulus to
match one of the events already stored through the previous training
sequences.
2
CA 02432751 2003-06-20
It is yet another object of the present invention, the capability of
generating
notification signals upon detecting the present of an alarming event.
BRIEF DESCRIPTION OF TIIE DRAWINGS
The features and advantages of the invention wili become apparent from the
following detailed description of the invention in which:
Figure 1 is an exemplary block diagram of a system that receives sound signals
and when detecting a matching within its own database generates a notification
signal.
Figure 2 is an exemplary block diagram showing the main components of the
system processing the acoustic signals described by the present invention.
Figure 3 is an exemplary block diagram showing one embodiment of the
invention where two or more devices are networked together such, that all of
them except
one will do the detection only, and upon'the detecting an alarm will trigger
themselves a
specific sound that will be detected by the main device that has connection to
the phone
line/Internet.
Figure 4 is an exemplary diagram showing one hundred samples of. a 3 kI-Iz
sinusoidal signal (free of noise) as it is after the sampling process.
Figure S is an exemplary diagram showing one hundred samples of a 3.3 kHz
sinusoidal signal (with white noise) as it is after the sampling process.
-,
J
CA 02432751 2003-06-20
DETAILED DESCRIPTION OF TFIE PREFERRED EMBODIMENTS
In the following description certain terminology is used to describe features
of the
present invention. For example an "alarm sound" is any acoustical
manifestation of a
phenomenon that someone might want to be notified on.
A "Control Logic" is any hardware/software instantiation of an algorithm
operating on
signals that are expressed as numbers. A "notification signal" is ar~.y
electronic signal that
can be activated for the purpose to inform a rwmote entity that the specified
alarm has
occurred.
First embodiment
Referring to Figure l, an illustrative alarm processing system is sh~wn, where
the
user inputs 100 allows the user to instruct the system about what sound alarms
to be
detected when analyzing the incoming sound 103, and how to submit the
notification
102. The system itself 101 communicates with the external world through a
microphone,
a network interface and a user interface.
Referring to Figure 2, an illustrative architecture instantiat~ng the
invention is
shown, where the Control Logic 204 is a combination of hardware and software
that
performs an analog to digital conversion of the sound captured by the
microphone 202,
calculates a digital signature and compare that signature with all the digital
signatures
already calculated and saved in the Memory 201. The communication between the
4
CA 02432751 2003-06-20
Memory and the Control Logic is performed via a digital bus 200. Upon
detecting a
matching between the two digital signatures, a notification signal will be
sent through the
communication channel 205 to the Network interface 206. The analog signal
generated
by the microphone is also sent to the Network Interface upon detecting a
matching
between two digital signatures.
Second embodiment
Referring to Figure 3, another embodiment of the present invention is shown.
In the case of a large area that is protected by an integrated alarm system as
invented
here, there will be a few devices 300, 302, 306 that are spatially distributed
and
implement the processing section, acting as transmitters, and instead of the
network
interface they will have a sound generating device such that when the alarming
sound 30I
occurs, a notification sound 303 will be generated. Upon receiving the
notification sound,
the receiver 304 that is just another instantiation of the invented device
will recognize the
notification sound and generate a notification signal 305 across the Network
Interface:
Referring to Figure 4, an illustrative representation of the waveform as
resulting
after the analog to digital conversion is shown. The waveform corresponds to a
sinusoidal
signal at 3 kHz, noise free.
Referring to Figure 5, an illustrative representation of a variation of the
signal
shown in Figure 4 is depicted. The waveform corresponds to a sinusoidal signal
3.3 kHz,
with half of the amplitude of the signal represented in Figure 4 and with a
white noise
added.
CA 02432751 2003-06-20
The present invention may be embodied in other specific forms without
departing
from its spirit of essential characteristics. The described embodiments are to
be
considered only as illustrative and not restrictive. The scope of the
invention, is
therefore, indicated by the appended claims rather than by the foregoing
description.
All changes that come within the meaning and range of equivalency of the
claims are
to be embraced within their scope.
DETAILED DESCt~IPTION OF THE INDENTION
In the particularly advantageous embodiment of the invention illustrated, the
interaction with the medium is performed by the microphone that converts the
acoustical stimuli 103 (as shown in Figure 1) into electrical signal, the user
inputs 100
and the network interface that creates the signal for the phone line/internet
102. These
are just the interfaces, and the whole system 101 also contains the analog to
digital
converter, the Memory and the Control Logic that were not depicted in Figure
1.
Upon the alarming sound is developing and the user is giving an appropriate
command for that alarm sound to be analyzed and stored into memory, the data
that
results after the analog to digital conversion is processed by the Control
Logic and a
digital signature of that signal is stored into the memory. It is not the
purpose of this
invention to claim the specific method of compression or how to generate the
digital
signature, and any compression method that gives good results can be employed.
For
illustration purposes only Figure 4 shows a waveform representing a 3 kHz
sinusoidal
signal, noise free, and Figure S represents a waveform of a 3.3 kHz sinusoidal
signal
6
CA 02432751 2003-06-20
half of the amplitude with noise added. The two signals can be considered as
generated by the same source or not (under different conditions), depending on
the
algorithm that calculates the digital signature through data compression and
the
algorithm that compares the digital signatures of each of them.
The same procedure that was employed to calculate and store one signature can
be
employed multiple times to calculate and store multiple alarm sounds.
Figure 2 illustrates the main blocks that allow the digital signatures to be
calculated
and compared. The microphone 202 produces the electrical signal 203 that is
processed by the Control Logic 204 and the results are stored into Memory 201
via
the signal bus 200. Also the Control Logic performs the comparison between the
already stored digital signatures and the digital signature of the signal
present at the
input of the microphone. When a matching is detected, a signal 205 is
generated
toward the Network Interface 206.
In one of the possible embodiments of this invention, for the purpose to deal
with
large spaces where the sound can be attenuated, a plurality of receivers 300,
302, 306
as described in Figure 3 can be deployed, where each receiver will have the
capability
of converting acoustic signals into electrical, calculate and compare the
digital
signature, but instead of driving a network interface, an acoustical signal
303 will be
generated and that signal will be later captured and processed by the unit 304
that will
drive the network interface through the signal 305. 13y doing this, a
plurality of
0
alarming sounds can be handled, each of them being possible to happen in
different
places across a given area of a house, office or more generic the area that is
intended
to be monitored for alarming sounds
7
