EP1932014A1

EP1932014A1 - System for tracking a mobile target

Info

Publication number: EP1932014A1
Application number: EP06806984A
Authority: EP
Inventors: Gordon Jones; Joe Spencer; Charalambos Lappas; Hui Mun Looe
Original assignee: University of Liverpool
Current assignee: University of Liverpool
Priority date: 2005-10-04
Filing date: 2006-10-04
Publication date: 2008-06-18
Also published as: GB0520187D0; WO2007039631A1

Abstract

A method and system for tracking a mobile target within a monitored space using at least three fixed directional detectors having respective responsitivities which vary with angular position around the detector and being positioned around a reference axis with their respective detection axes angularly spaced apart and inclined towards said monitored space such that the detectors have angularly overlapping response characteristics and a combined non-uniform resonsivity profile. A signal is emitted at a source on the mobile target so that the at least three detectors provide respective electrical output signals (R,G,B). Chromatic processing is effected to transform the output signals (R, G, B) into at least first and second chomatic parameters (Hs, Ls) whose values are dependent upon the position of the target in said monitored space and the position of the target is determined based on the values of the chromatic parameters (Hs, Ls).

Description

DESCRIPTION

SYSTEM FOR TRACKING A MOBILE TARGET

The present invention is concerned with a system and method for tracking a mobile target.

The present invention makes use of known non-orthogonal processing

techniques, the principles of which are described briefly hereinafter by way of background information.

Non-orthogonal processing techniques make use of non-orthogonal response characteristics in signal processors. A "non-orthogonal" system is one wherein the responses of processors, e.g. detectors, in a signal domain (e.g. optical wavelength)

overlap, as illustrated in Fig. 1 of the accompanying drawings. As evident from Fig. 1 , as a result of the overlapping in the signal tails, the outputs of the detectors are cross-correlated, yielding higher sensitivity to signals in the tails.

In principle, the signal processors used in a given non-orthogonal monitoring system will be responsive in a particular signal domain. The signal domain may in

principle be any of a plurality of conventional signal domains, including optical and radio, each addressed in the frequency (wavelength) or time domains. Additionally,

it has been established that other domains such as spatial location, mass, and non-

orthogonality between specific parameters (e.g. pressure and temperature etc) plus

combinations of large numbers of sensor types can be accommodated. However, the

invention described herein is based on the situation where the monitoring signal

domain is essentially acoustic and monitoring is achieved by the use of so called " chromaticity processing", the principles of which are now well established in the an.

Explanations of the details of "chromatic processing" can be found in many patent

specifications and papers published over the last fifteen years or more to which

reference is hereby directed, for example:-

EPO 214845 based on GB 8522202 filed by The University of Liverpool in

1985:

"The Gabor transform basis of chromatic monitoring" by G.R. Jones, P. C. Russell, A. Vourdas, J. Cosgrave, L. Stergioulas and R. Haber, Measurement Science

Technology 11(2000) 489-498;

"Position location of a filamentary arc using a tristimulus chromatic

technique" by Y. Yokomizu, J. W. Spencer and G.R. Jones, J.Phys. D: Appl Phys 31 (1998) 3373-3382;

"Chromatic Modulation Based Acoustic Analysis Technique for In-Process

Monitoring of Laser Materials Processing" by J.M. Lee, K.G. Watkins, W.M. Steen,

P.C. Russell and G.R. Jones, Journal of Laser Applications 11 199-205 (1999).

"Parameter monitoring using neural-network processed chromaticity" by S.

Ahmed, P. Russell, P. Lisboa, G.R. Jones, I EEE Proceedings - Science Meas

TechnoL VoI 144, No 6. Nov 1997.

As explained in the quoted references, "chromaticity processing" is the name

given to the application of sets of non-orthogonal weighted integrals to signals

distributed across a measurement range and the subsequent transformation of the

integral quantities obtained to give parameters summarising certain characteristics of

the distribution. The name derives from the origin of the method in broadband optics and colour science, where the distribution to which it is applied is that of light

intensity across the optical spectrum. However, it is applicable to measurements of

any quantity distributed across another variable (for example, acoustic intensity with

frequency or temperature with spatial position). Where N integral weightings take the form of Gaussian curves (Fig. 2) the quantities derived in the first part of the

process are the values of N basic terms of a Gabor expansion of the original signal.

It has also been shown that this process is maximally information preserving for the general signal and that useful information retention with high robustness to noise is obtainable with as few as three Gaussian integrals.

In the optical domain, an approximation to these three Gaussian base integrals

can be provided by the wavelength response of the sensor elements (e.g. colour photo detectors in CCD cameras). This is known as a tristimulus sensor system.

Observations may therefore be represented as data points in a colour space, the most

straightforward of which is a Cartesian colour cube having an axis for each of the three sensor elements. The three co-ordinates of a point therefore give a separate

measure of each of the familial" red, green and blue components of visible light.

Thus, where the original data is a visible spectrum, these axes correspond to the red,

green and blue components of a colour and such colour terminology is often applied

by analogy where other distribution variables and measurements are involved to aid interpretation.

The second stage in chromatic processing (which ma)^r, in some

circumstances, be omitted, or, where there are only a few discrete values of the

distribution variable, be used on its own) is the transformation of the Cartesian colour - A - space into a space referenced by a new set of parameters. These new parameters are

formed by the combination of the tristimulus parameters according to a set of

formulae that describe the transformation. Several such transformations have been

established in colour science, but one in particular has been found to be especially

useful for the contribution that it makes to operator interpretability of information

through its partitioning into components of distinct character. This is the transformation to HLS (Hue, Lightness, Saturation) space. By way of example only, the transformation can be

/ 60(G-B) /highest of R,G,B-lowest of R₅G₅B, if highest of R₅G₅B=R f 60(2+(B-R)) /highest of (R,G,B)-lowest of (R,G,B)), H = 1 if highest of R₅G₃B=G . -(1)

/ 60(4+(R-G)) /(highest of (R₅G₅B)-Io west of (R₅G₅B)), I if highest of R₅G₃B=B

R + G + B (2)

3 highest of (R.G.B)-lowest of (R.G.B) (3) highest of R,G,B+lowest of (R₃G₅B)

where R₅ G and B are the measured instantaneous values of the red, green and

blue parameters of the Cartesian colour space, and H, L and S are the hue, lightness

and saturation components of the new space.

Hue is specified as an angle (given in degrees by the above formula) and the

lightness and saturation parameters range from 0 to 1, giving a cylindrical polar space

of unit radius and axial extent (Fig. 3). These parameters partition the information

acquired such that lightness corresponds to the nominal amplitude of the original

measurements summed across the rεnse of their distribution variable, saturation indicates the degree to which the measurements are spread throughout the range of

the distribution and hue corresponds to an effectπ e value of the distribution variable

about which the measurements are spread. The parameter names reflect the

interpretation of these characteristics familiar from colour perception. Where the measurements are of quantities other than visible light, physically analogous and

informatically identical (but for some small departure of our colour receptors from a Gaussian response) processing provides an intuitive assimilation of the information represented.

Chromatieity processing has relied conventionally upon the non-orthogonality of plural optical detectors for classifying detected signals. In this connection, colour (which is a human perception) may be regarded as a special case of chromatieity, whereas chromatieity may itself be regarded as a special case within the more general

area of non-orthogonal signals discrimination.

Each detected signal has a special signature which may be classified by N

defining parameters. In general such signatures form highly non-linearly related sets

requiring the need for at least N=3 defining parameters for classification in signal

space (tri-stimulus processing). (The use of N=2 parameters (distimulus) constitutes

a linear approximation in two dimensional signal space).

By way of example, if it is assumed that all signals are Gaussian distributions

of variable signal strength with respect to the signal domain (e.g. wavelength,

frequency, time etc), classes of signals are then unambiguously defined by onl} N=3

parameters corresponding to (figs. 4a and 4b):-

• Signal amplitude (or power content) (L) • Location of the peak value in signal parameter space (H)

• Signal half width (S)

If the need for all signals to be Gaussian in nature is relaxed, then each signal may be allocated to one only of a class governed by a mother Gaussian. This

provides a substantial but not absolute signal discrimination means through the use

of only three detectors (R, G, B) to yield three functions H,L,S. This forms the basis

of chromatic discrimination: if the forms of the R, G, B detectors correspond to the responsivities of the human eye, the N=3 chromaticity degenerates into the special case of colour. H,L,S are then N the Hue, Lightness: and Saturation of colour science as described above.

Extension of the aforegoing technique to the use of N > 3 parameters leads

to a subdivision of each mother Gaussian class into additional non-Gaussian classes (see Fig. 4b). By way of an example, N=4 may define the degree of

asymmetric deviation (Skewness) from a Gaussian distribution (see Fig. 4c) i.e.

each Gaussian class N subdivides into several asymmetric Gaussians

x ∑ n_s , with x being determined by the signal processor discrimination. Furthermore s=l

an extension to N = 5 parameters enables the degree of Kurtosis of the Gaussian

distribution to be determined (see Fig. 4d) leading to a further subdivision of each

asymmetric Gaussian class into £ h_k subclasses.

K=I

As mentioned in several places hereinbefore, the signal processors used in a - i - given non-orthogonal monitoring system will be responsive in a particular domain.

The aforegoing background discussion has been made on the basis of the optical

domain. However, as stated, it has been appreciated that the domain need not in

principle be optical and indeed preferred embodiments of the present invention makes use of measurements made in the spatial acoustic domain.

The important point to be noted here is that, for spatially dependant input

signals, eg acoustic signals, exactly the same processing approach and assumptions can be made. Thus, "chromatic processing" can be applied to spatially dependant input signals in the acoustic domain in a precisely analogous manner. The same

applies to signals in the rf and infra-red domains, for example, as well as the optical domain. Detector signals can therefore in all such cases still be given the identifying letter R, G, B leading to "chromatic parameters" H, L and S, etc.

In our co-pending PCT/GB2005/0 1044 based on GB 0407274.0 we describe

a system for tracking a mobile target, comprising at least three fixed directional

acoustic detectors and an acoustic source as the mobile target, the detectors providing

corresponding respective electrical output signals which are arranged to be

chromatically processed whereby to yield chromatic parameters whose values are

dependent upon the target position. In that case, the at least three fixed directional

acoustic detectors were disposed with their angularly spaced apart axes extending in

a common (horizontal) plane.

An object of the present invention is to provide an improved system and

method for identifying the location and/or status of one or more objects or points in

space making use of the aforegoing "chromatic processing^" technique. - o -

In accordance with the present invention there is provided a method of

tracking a mobile target within a monitored space using at least three fixed directional

detectors having respective responsivities which vary with angular position around

the detector and being positioned around a reference axis with their respective

detection axes angularly spaced apart and inclined towards said monitored space such

that the detectors have angularly overlapping response characteristics and a combined non-uniform responsivity profile:

the method comprising: emitting a signal at a source on the mobile target so that the at least three detectors provide respective electrical output signals (R, G, B);

effecting chromatic processing to transform the output signals (R, G, B) into at least first and second chromatic parameters (H₅, L_s) whose values are dependent upon the position of the target in said monitored space; and

determining the position of the target based on the values of the chromatic parameters (H_s, L_s).

Preferably, the method includes obtaining and establishing a store of values

of angular position (θ) of the source relative to a reference direction based upon

values of the first chromatic parameter (H,) and obtaining and storing values of radial

distance (r) of the source from said reference axis based upon values of the second

chromatic parameters (L₅), modified in dependence upon values of the first chromatic

parameter (H₅) to compensate for said collective non-uniform responsivity profile of

the detectors; and

comparing said stored values of angular position (θ) and radial distance (τ) with values of the chromatic parameters (H, L) calculated from measured values of

the detector output signals (R, G, B) to establish the instantaneous position of the

target.

In some preferred embodiments, the output signals (R, G, B) are transformed into three chromatic parameters (H, L, S) from the equations:-

60(G-B) /(max(R,G,B)-min(R,G,B)), if max(R,G,B)=R H - 60(2+(B-R)) /(max(R,G,B)-min(R,G,B)), if max (R,G,B)=G (1) 60(4+(R-G)) /(max(R,G,B)-min(R,G,B)), if max (R_;G,B)=B

Q maxCR. G. BVmJn(R. G. B^ max(R, G, B)+min(R, G, B) (3)

where R, G and B are the measured values.

In other preferred embodiments, the transformations can be obtained from:

r=R-min(RGB) g=G-min(RGB) (4 )

b=B-min(RGB)

Lightness (L)= {^5&Z^≡^}≡^ 16) ¹ 2 i

Saturation Advantageously, the method Includes the use of a fourth fixed, directional

detector having a respoiisivity which varies with angular position around the detector

and which produces an output signal (F), the fourth detector being disposed with its detection axis inclined towards said monitored space at an angle different to that at the first mentioned three detectors.

Preferably, the response characteristic of the fourth detector overlaps the

response characteristics of all three of said first mentioned detectors.

An additional chromatic parameter (S_F) can be formed from the output signal (R, G, B, F) of the four detectors whose value is dependent upon the height of the target in said monitored space.

In some preferred embodiments, the additional chromatic parameter (S_F) is defined by:

S_F = greatest of (R₅ G, B) - F greatest of R, G, B + F

The values of the further chromatic parameter (S_F) can be calibrated against

measured values of the height of the target in said monitored space and stored, for

example in a look-up table.

Advantageously, the values of the further chromatic parameter (S_F) are

modified by a second calibration factor (C) dependent on the value of the first

chromatic factor (H_s) to compensate for the variability of the amplitude of the further

chromatic parameter fS_F) with the value of θ.

In accordance w ith a further aspect of the in\ ention. there is provided a

5} stem for tracking the position of a mobile target within a monitored space. comprising:- at least three fixed directional detectors having respective responsivities

which vary with angular position around a reference axis with their respective

that the detectors have angularly overlapping response characteristics and a combined

non-uniform responsivity profile; a source (S₀) adapted to be carried by the mobile target and to emit a source signal; the at least three detectors providing respective electrical output signals (R,

G, B); means for effecting chromatic processing to transform the output signals (R, G, B) into at least first and second chromatic parameters (H_s, L_s) whose values are dependent upon the position of the target in said monitored space; and

means for obtaining values of the first and second chromatic parameter (H_s,

L_s) calibrated such that the value of the angular position (θ) of the source from a

reference direction corresponds to the value of the first chromatic parameters (H_s)

and the value of the distance (r) of the source from the reference axis corresponds to

the value of the second chromatic parameters (L_s), modified in dependence upon the

value of the first chromatic parameter (H₅) such as to compensate for said collective

non-uniform responsivity profile of the detectors.

In accordance with a first embodiment of the present invention there is

provided a method of tracking a mobile target using at least three fixed directional

detectors (M₁, M-, M₃) having respective responsivities which vary with angular position around the detector and being disposed in a common plane at or uniformly

spaced about a reference axis, with their respective detection axes angularly spaced

apart and inclined relative to said common plane such that the detectors (M₁, M,, M-)

have angularly overlapping response characteristics, the method comprismg:-

emitting a signal at a source on the mobile target so that the at least three

detectors provide respective electrical output signals (R_s, G_s, B_s);

effecting chromatic processing to perform a plurality of cross-correlations on

the output signals (R_s, G_s, B_s), to transform the latter signals into at least first and

second chromatic parameters (H_s, L_s) whose values are dependent upon position of

the target relative to the reference axis;

calibrating and storing the first and second chromatic parameters (H_s, L₅) such

that the value of the angular position (θ) of the source in said horizontal plane from

a reference direction through said reference axis corresponds to the value of the first

chromatic parameter (H_s), and the value of the radial position (r) of the source from

the reference axis corresponds to the value of the second chromatic parameter (L_s),

modified by a calibration factor (A) dependent on the value of first chromatic

parameter (H_s) such as to compensate for the collective non-uniform responsivity

profile of the detectors; and

comparing the calibrated stored values of radial position (r) and angular

position (θ) with values of the chromatic parameters (H, L). calculated from

measured values of R₅ G. B to establish the instantaneous co-ordinates of the target.

Advantageous!} , the method comprises using at least one further fixed,

directional, detector (M₄; which is disposed so that its axis is inclined from said common plane at an angle different to that of the first mentioned three detectors (M₁,

M₂, M₃) and with its response characteristic overlapping the response characteristics

of the latter detectors (M₁, M₂, M₃), and wherein an additional chromatic parameter

is formed from the output signals of the four detectors whose value (S_F) is dependent upon the distance of the target from said common plane.

In some preferred embodiments, the additional chromatic parameter (S_F) is defined by:-

greatest of(R. G. B) - F

^F greatest of (R, G, B) + F

Values of S_F, preferably modified by a second calibration factor (C) dependent on the value of the first chromatic parameter (H_s), are calibrated against measured values of the distance (z) of the source from said common plane and stored, for example in a look-up table.

In some cases, the fourth detector (M₄) is disposed with its axis passing through said reference point.

Advantageously, the fourth detector (M₄) is disposed with its axis

perpendicular to said common plane, preferably vertically downwardly.

Preferably, the detectors are acoustic detectors and the source emits an

acoustic frequency signal which can be ultrasonic.

Preferably, the acoustic signal comprises a single frequency.

In accordance with another aspect of the invention there is provided a mobile

target tracking system comprising:-

at least three fixed directional detectors (M,, M₂, M₃) having respective non- uniform frequency responsivities which vary with angular position around the

detector and being disposed in a common plane at or uniformly spaced about a

reference axis, with their detection axes angularly spaced apart and inclined relative

to said common plane such that the detectors (M₁, M₂, M₃) have angularly

overlapping response characteristics; a source (S₀) adapted to be carried by the mobile target and to emit a source

signal; the at least three detectors (M₁, M₂, M₃) providing respective electrical output

signals (R_s, G₈, B_s); means for effecting chromatic processing to perform a plurality of cross- correlations on the output signals (R_s, G_s, B_s), to transform the latter signals into at least first and second chromatic parameters (H_s, L₅) whose values are dependent upon the horizontal position of the target relative to the reference axis, and;

means for obtaining values of the angular position (θ) of the source in said

horizontal plane from a reference direction through said reference axis based upon

the value of the first chromatic parameter (H₈), and means for obtaining the value of

the radial position (r) of the source in said horizontal plane from the reference axis

based upon the value of the second chromatic parameter (L₅), modified by a

calibration factor (A) dependent on the value of first chromatic parameter (H_s) such

as to compensate for the collective non-uniform responsivity profile of the detectors.

Preferably, there is a single acoustic source on the target, although additional

sources orientated with respect to each other mav be carried on the target.

Thus, the measurement domain for the present invention is spatial position addressed preferably via an acoustical detection system. Whereas in the optical

domain the chromatic processor has a response which is a function of optical

wavelength, in the present case the chromatic processor, which is preferably

acoustical in nature, has a response which is a function of angular position. Thus the

approach involves the use of three or more such processors, eg acoustical ultrasonic

processors, having non-orthogonal responses in three dimensional space.

In one preferred arrangement, said three fixed, directional detectors comprise

respective microphones disposed in a star arrangement with their axes mutually

spaced by 120°.

In some embodiments using a star array of microphones, three additional

detectors are employed in a delta formation at the periphery of the monitored space,

inclined at 120° to each other for enhancing the discrimination ability of the system.

Thus, both star and delta arrangements may be deployed together to provide

greater enhancement and forming an N=6 chromatic system.

In another preferred arrangement, said three fixed, directional detectors

comprise respective microphones disposed in a delta arrangement at the periphery of

a measuring space with their axes mutually spaced by 60° .

Preferably, each acoustic source is an ultrasonic source.

In some embodiments, a second stage space chromatic processing is

performed on said chromatic parameters as a function of time. wherebv to generate

information as to the movement of the target.

In accordance with yet another aspect of the present invention there is

provided a system which enables a mobile target to determine its location, comprising at least three fixed chromatically directional sources and a detector on the mobile

target. Each source transmits a coded signal (e.g. frequency, time sequence, etc.) so

producing three electrical output signals by the mobile detector which are arranged

to be chromatically processed to yield chromatic parameters whose values depend

upon the position of the mobile target.

It is emphasised that in techniques described herein, which are preferably acoustical, movement can be detected from a known acoustical signal emitted by an object to be tracked, and not by acoustic signals produced by points in the environment. Thus, the object itself is effectively being monitored. The chromatic

addressing is undertaken in the spatial domain by processors whose acoustical responses vary with angular position, i.e. non-orthogonality between the angular responses of acoustical detectors each having the same acoustical frequency responsivity.

The invention is described further hereinafter, by way of example only, with reference to the accompanying drawings, in which:

Fig. 1 illustrates the responses of three detectors having overlapping response characteristics;

Fig. 2 shows examples of Gaussian response curves superimposed upon a signal:

Fig. 3 shows H, L and S in cylindrical polar space;

Fig. 4a shows how Gaussian signals are unambiguously defined b> H. L and

S values:

Fig. 4b shows how other signals are defined as the Gaussian family to which they belong;

Fig. 4c shows how the use of four processors gives a measure of skewness;

Fig. 4d shows how the use of five processors gives a measure of kurtosis;

Fig. 5 illustrates diagrammatically the disposition of acoustic detectors in one embodiment of the present invention;

Fig. 6 shows a receiver unit comprising four acoustic detectors attached to a ceiling surface;

Fig. 7 shows how the 3-D location of a tag carrying an acoustic source can be derived through H, L, S transformations of acoustic detector output values R, G₅ B;

Figs. 8a and 8b show typical relationships between the source position coordinates r, θ and the chromatic parameters H, L for the tre-foil cluster M₁, M,, M₃;

Figs. 9a and 9b show the responsivities of two different acoustic detectors;

Fig. 10 shows the H-L characteristic for the detector having the response characteristic of Fig. 9a;

Fig. 11 shows a plan view of an environment being monitored by a four-

detector system;

Fig. 12 shows how the 3-D location of a tag carrying an acoustic source can

be derived through H, L. S transformations of acoustic detector output values R, G.

B and F:

Fig. 13 shows how the S_F parameter varies with source height;

Fig. 14 shows signals obtained when monitoring two separate acoustic

sources;

Figs. 15 and 16 show acoustic detectors arranged in star and delta arrangements, respectively;

Fig. 17 shows the use of further acoustic detectors in a delta array around a

tie-foil star detector cluster;

Fig. 18 illustrates "second generation" processing; and

Fig. 19 shows examples of H-L, H-S chromatic diagrams for a tag at three different positions.

Figures 5 and 6 illustrate an arrangement of acoustic detectors for a system in accordance with one embodiment of the present invention for monitoring the position of a target within a 3 -dimensional space, such as a room. In this

embodiment, three directional microphones M₁, M₂, M₃, i.e. microphones whose response characteristic are angularly non-uniform and therefore, more responsive in some angular directions than others, are disposed as part of a receiver unit 10 in a trefoil arrangement with their axes orientated at 120° to each other from a common reference point P and with 45° between their axis and the horizontal plane containing

the reference point P. In a typical practical embodiment, the horizontal plane would

normally correspond to a horizontal ceiling surface of a room/space to be monitored

by the system (or be disposed parallel to and below said ceiling surface).

Preferably, there is a fourth acoustic detector, for example in the form of a

further directional microphone M₄, whose detector axis passes through the reference

point P and is directed downwardly from said horizontal plane towards the space to

be monitored. In many environments to be monitored, for example substantiall}-

square room spaces, the fourth microphone is preferably directed downwardly at right

angles to the horizontal ceiling plane but in some cases, for example if the room space is highly rectangular (narrow) it can be advantageous to angle the fourth

microphone away from the vertical in a direction generally along the length of the

room space.

Normally, the receiver unit 10 carrying the tre-foil microphone arrangement and the fourth detector M₄ would be disposed at or near the centre of the ceiling of

the 3-dimensional space to be monitored.

Figure 7 shows how the measured values R, G, B from the three microphones

M₁, M₂, M₃ can be used, through H, L, S transformations of the R, G, B values, to identify the location of a target disposed/moving in a two-dimensional plane at a fixed height below the two microphones in terms of its two spatial co-ordinates r, θ (radial distance and horizontal angle in relation to the reference point P).

hi a preferred embodiment, the target carries a source of acoustic signal which preferably generates a pure sinusoidal signal, for example of 4OkHz. hi this case, the detectors M₁, M₂, M₃ are preferably ultrasonic detectors operating at a nominal resonant frequency of 40IcHz. The acoustic source may, for example, be in the form

of a tag, carried by a person whose position is to be monitored, and which is

orientated normally with its axis directed generally upwards towards the ceiling.

As shown by Fig. 7, the microphones M₁, M₂. M₃ produce output signals R,

G, B respectively which are amplified and rectified to give discrete analogue values

of R, G, B at any instant. In practice, the latter values are digitized and passed to a

suitable computer or computing chip (not shown) which performs the signal

manipulations and calculations as described hereinafter.

Tag position is located via r. θ (and possibly z) co-ordinates using the three star orientated ultrasonic receivers M₁, M₂, M₃ shown in Figs. 5 and 6, the outputs of

the three receivers forming the R, G. B channels of a chromatic monitoring system

to yield chromatic parameters Hue (H), Lightness (L) and Saturation (S). L is

dominated by radial position (r) and H by azimuth position (θ). The fourth channel

M₄, when used, assists the 3-D positioning of the tag (z co-ordinate) as well as the

elimination of signal reflections and tag inclination effects.

Referring again to Fig. 7, this shows the process whereby the spatial coordinates (r, θ) of a source are determined from the outputs (R, G, B) of the three acoustic detectors M₁, M₂, M₃ via the technique of chromatic transformation to the parameters referred to for convenience as H, L, S (Hue, Lightness, Saturation).

The R, G, B output signals are converted to H, L, S chromatic parameters using well known algorithms in the chromatic processing art. For example, in a basic embodiment the parameters can be those quoted hereinbefore, namely

_T, . . . _{+ +} , _ττ 60 (G-B) lhis is interp ¹ reted as: H = h. i.gh. est . of ^ R_n. ^ G and _J B _n-smal „lest , o _ff R_n, ^ G and , B_n

when the highest of R, G and B is R

G and B-smallest of R, G and B

when the hi -ogh- est of R. G and B is G

_{H =} 60 ⁽4^fR-G) ) highest of R. G and B-smallest of R. G and B when the highest of R, G and B is B

R + G + B

L = -(2)

highest ofR, G and B -lowest of R. G and B -(3)

Q highest of R, G and B+lowest of R, G and B

However, in a preferred embodiment, the equations for calculating the H, L and S values are obtained via> r=R-min(RGB) i.e (r=R-lowest of R, G and B) g=G-min(RGB) i.e (g=G-lowest of R, G and B) (4) b=B-min(RGB) i.e b=B-lowest of R, G and B

(highest of R. G and B-lowest of R. G and B)

Lightness L = 2 M

vasxiRGE) - ΏMRGB)

Saturation S =

VOSSiRGE) * mm(RGB)

(highest of R. G and B-lowest of R. G and B) __(3) highest of R, G and B÷lowest of R, G and B These three values Hue, Lightness and Saturation (HLS) characterise the signal and represent some of its features.

The system is calibrated by recording the values of H, L, S for individual signals received from an acoustic source placed sequentially at a number of spatial positions in 2-dimensional space around the tre-foil cluster of detectors (M₁, M₂, M₃),

these positions being chosen with respect to (r, θ) to be representative of the different signal effects which can occur in a given environment.

The calibration values (H_c, L_c, S₀), are stored as a look up table with respect to r, θ in either the aforementioned computer or a programmable micro chip.

Alternatively, for detectors with appropriate angular responsivities, analytical expressions relating r, θ to H, L, S may be formed e.g.

θ =f, (H) (8)

1 r = - [In [LoZL)- /, (H)] (9) n

As indicated in Figure 7, the angular position θ corresponds uniquely to

values of H, the precise relationship being determined by the form of the detector

responsivity, as described further hereinafter. The radial position r is primarily a

function of L but moderated by a calibration factor A (figure 7) determined from the

measured H value with the particular value of L as given by the look up table. The

modified (new) L value yields the r position from the look up table. The calibration factor is required to take into account the collective non-uniform responsivity profile

of the detectors (resulting in the "butterfly" effect on the profile of the detector outputs as shown in Figs 8b and 10) arising from the fact that practical detectors yield

different L responses at different (hue) angles for the same distance from the

reference point P (see Figs. 5 and 6 discussed hereinafter) because of their nonuniform response in different directions.

The third chromatic parameter (S) has a unique value corresponding to the calibrated values H_c, L_c. However, if unexpected reflections or other aberrations

occur then the measured S value will differ from that determined from the look up table corresponding to the particular H, L values. Consequently, a threshold may be set beyond which values of a given H, L pair, the S value differs from the calibrated S value to indicate that the position measurement (r, θ) may be less accurate due to such unexpected aberrations.

Figure 8 shows, by way of an example, typical relationships between the

source position co-ordinates r, θ and chromatic parameters H₃ L for a tre-foil cluster M_], M₂, M₃ having the measured responsivities with respect to a radial axis aligned

to the maximum responsivity of M₁ (Fig. 9a) and corresponding to θ=0 (Fig. 8b).

Figure 8a shows the variation of chromatic parameters L with the radial

position r of a source for two different angular positions θ corresponding firstly to the

angular direction of maximum sensitivity of detector M₁ (i.e. θ=0 figure 8b, figure

9a) and secondly to the angular direction of crossover between the responsivities of

detectors M₁ and M₂ (i.e. θ=60° figure 8b, figure 9a). The L variations with r at both

angles are of a similar form but the amplitude of L for θ=0 is greater than at θ=όθ°

(so satisfying the H, L equations).

Figure 8b shows an L:H polar diaεram with zwo examples of a locus of a source moving around the tre-foil cluster (M₁, M₂, M₃) at a constant radius. The

continuous "butterfly" curve corresponds to points on a circle of radius Im, and the

dashed curve to a circle of radius 2rn. Thus, source at a distance of Im from M₁, M₂, M₃ would yield L=OdB at 0° but -5dB at 60°.

Figure 9 shows examples of the angular responsivities with two different

types of detectors, each with three detectors (M_j, M₂, M₃) deployed non-orthogonally with respect to each other, the angular responsivities in each case being shown with a linear angle (θ) scale, consistent with the deployment of chromatic detectors of polychromatic light used in the known chromatic processing of wavelength responsive detectors.

The two characteristics (figure 9a and 9b) show how the non-orthogonality (overlap) of the detectors can differ for two types of detectors, the angular extent of the overlap being greater for the figure 9a detector compared with the figure 9b detector.

Because of the circular symmetry of the detector cluster geometry with respect

to angular position θ, the responsivity of each detector is non-orthogonal with respect

to the other two detectors whereas in conventional chromatic deployments (eg

colourimetry) which are linear in nature (eg Fig. 1), it is only the central detector

which is non-orthogonal with the other two detectors.

The nature of the non-orthogonality of the detectors affects the transformation

of the spatial co-ordinates (r, θ) into the chromatic co-ordinates (H. L). Thus,

whereas the detector system of figure 9b leads to the H-L characteristic shown on

figure 8b. the detector svstem of figure 9a leads to the different H-L characteristic shown on figure 10.

Consequently, the relative sensitivity of the monitoring system on various

regions of r. θ can be tuned by varying the detectors' responsivities as illustrated by

figures 8, 9 and 10.

When a person carrying a sound source tag changes his position in the monitored room, the R, G and B values (from M₁, M₂, and M₃) change accordingly. The values are input into the PC through an A/D interface for post processing and

Hue, Lightness and Saturation values are calculated. Figures 19a, 19b, 19c are the

resulting computer screen display and show the person in various positions.

Figures 5 and 6 show the deployment of a fourth detector (M₄) along with the tristimulus detectors M₁, M₂, M₃. Whereas M₁, M₂, M₃ are deployed at 120° with respect to each other and 45° between each detector axis and the horizontal, the fourth detector M₄ is situated central to M₁, M₂, M₃ and vertically (i.e. 45° to the axes OfM₁,

M₂, M₃). Consequently, this arrangement constitutes a four detector system (N=4)

deployed in three dimensions so that the non-orthogonality (overlap) of the detector responsivities becomes more complicated than if there are just three detectors with

their detector axes located in a common (horizontal) plane.

Figure 11 shows a plan view of an environment being monitored by the four

detector system (approximate floor area 6m x 3m). The area cohered by the

tristimulus detectors M_j. M₂, M- is shown as L and the area covered b} the fourth

detector M_^ is shown as the circle t. The overlap of M₁. M₂. M₃ and M_ is apparent

on figure 1 1. [It will be appreciated from Fig. 11 that with the gain settings of the

detectors being used, onl> part of the room is covered. In order to cover ail of the room it would be necessary to increase further the gain of all of the detectors.]

Figure 12 shows the procedure for processing the four detector outputs. The

procedure for evaluating the spatial co-ordinates r, θ from L, H is identical to that for the three detector system of figure 7. The chromatic parameter S determined from

R, G, B is calibrated by experiment to be a function of the inclination of the source. Tests show that the relationship between source inclination and S_RGB depends upon

the angular position θ of the source so that a correction factor for θ (calibration factor B) needs to be incorporated (figure 12). The output of the fourth detector, F₃ is used

to form an additional chromatic parameter S_F which is defined by max (R.G.BVF (T)

^F max (R,G,B)+F

S_F is calibrated against the height position of the source by means of the outputs R, G, B, F when the source is placed at a number of different radial (r), angular (θ) and height (z) positions. Chromatic H, L and S_F values are derived from the R, G, B, F outputs using the algorithms defined by equations (3), (4), (5), (6) and

(7). These tests show that H and L are independent of the height position of the

source but that S_F varies with the height as shown on figure 13. Figure 13 also shows

that the S_F :source height characteristic remains of the same form at different angular

positions (θ) but the S_F amplitude is displaced by an amount dependent upon θ. The

raw S_F value is therefore modified by a calibration factor c (figure 12) whose value

is determined from the value of θ to yield z. Look up tables based on these

calibrations are used for determining the source position co-ordinates (r, θ, zj from

": *- ^■> -'RGB- "F- An advantage of the present system is that it can be adapted for monitoring

a plurality of acoustic sources simultaneously, for example on respective tags carried

by several people within the monitored space, as illustrated by figures 14a to 14c.

Figure 14a shows the form of signals produced by each of two sources. Each consist of a double burst of 4OkHz, each of duration t_a and separated by a time

interval I₁. With several sources, the duration t_d is equal but the separation interval t_j is different for each source. In the illustrated case, the signal bursts in the left-hand

case are separated by lms and in the right-hand case are separated by 2ms.

Each detector M₁, M₂, M₃ receives the double burst signal figure 14b from each source. Each burst is converted into a pulse and the time interval between pulses in a pair used to identify the source.

Figure 14c shows the signals from two different sources detected by one detector. Signal collision problems have been shown to be rare. By examining the resultant envelope of the received signal, it is possible to identify each tag.

Thus, there has been described above examples of spatial, non-orthogonal

detection systems based upon a tristimulus acoustic technique which is used for the

purpose of tracking one or more targets within a monitored 3 -dimensional space.

The described embodiments use three directional microphones M₁. M-,, M₃

in the case of Fig. 7 and four M₁, M₂, M₃, M₄ in the case of Fig. 12, together with one

or more acoustic sources which are arranged to be carried by respective mobile

targets. Thus, the or each source is itself mobile in that it moves with the associated

target.

The directionality of the microphones M₁, M,, M₃ M₄ is arranged to provide - Zo - the non-orthogonality for space discrimination required to enable the performance of

the chromatic processing techniques described.

In the embodiments described thus far. the three microphones M₁. M-, M-, are

clustered together but arranged to be directed at 120° to each others' axes so as to form a star arrangement, as shown in Figs. 1 and 15. However, in other

embodiments, similar microphones M₄, M₅, M₆ can be arranged in a delta

configuration, as shown, for example, in Fig. 16.

Still further embodiments can utilise 3 < N < 6 detectors in different forms.

For example, in one further embodiment, three additional detectors M₇, M₈, M₉ (ie N=6) may be employed in delta formation at the periphery of the detection volume to be monitored and relatively inclined at 120° in the horizontal plane, as

shown in Fig. 17. Such an embodiment may be used to enhance the discrimination ability of the system, for example with respect to reflections, scattering etc. of the ultrasonic signal from the source by artifacts within the monitored volume. The

outputs R_D G_D, B_Dfrom the three additional detectors M₇, M₈, M₉ may, for example, be processed to yield a further three chromatic parameters H₀, S_D, L_D which may be

cross-correlated with H_s, S_s, L_s.

The technique described thus far constitutes a first stage or "first generation"

chromatic processing based in the acoustic domain.

A second stage or "second generation" chromatic processing based in the

spatial domain (e.g. position within a space) may be applied to the first generation,

acoustic chromatic co-ordinates . In this case, the position of the source S₀ forms the

horizontal axis and the time duration of a chromatic disturbance of the source S₀ at a particular position forms the vertical axis. The time duration/position graph forms

a signal graph (Figure 18b) which is addressed by three non-orthogonal chromatic

processors (R_p, G_p, B_p) in the position (SPATIAL) domain.

Spatial chromatic parameters (e.g. H_p, G_p. B_p) are evaluated from the outputs

of R_p. G_p, B_p at various time instants and can be displayed on H_p - S_p, H_p - L_p polar

diagrams.

Alternatively, and preferably, H_p, L_p, S_p may each be displayed as a function of time (Figure 18d).

In this manifestation, H_p(t) represents the position of the acoustical source within the monitored volume as a function of time; L_p(t) represents the time duration for which the acoustic source remained located continuously at each location (Figure 18d).

Consequently the movement of the target carrying the source S₀ within the

monitored space may be tracked via the H_p(t) graph and the stationarity of the target- determined from the L_p:t graph as the increasing magnitude of Lp.

The second generation processing can thus be used in conjunction with the

H_p, L_p, S_p to yield quantifiable movement patterns.

If the basic delta pattern of detectors is used in place of the above described

star arrangement, the results are essentially the same in principle, with the angular

position θ and radial position again being established by algorithmic manipulation

of the H, L, S values.

It is possible that alternative systems can be employed in which the source and

microphones are interchanged, i.e. the target carries a microphone and there is a star or delta arrangement of acoustic signals. Signal sources can be arranged to be

directional and to be encoded, e.g. different frequency or time sequenced, so as to be effectively non-orthogonal and thus enable chromatic processing as described herein

to achieve similar results. However, in this case, the information regarding the target positions would be obtained at the target itself and not at the stationary transmitters

so that the system would be enabling the target to determine its own location and not for the position to be tracked remotely.

One application for the invention is in monitoring of individuals in institutions such as care homes. In this context information about a person's movements can be used to give an indication about their well being, for example. In

some such embodiments it may be desirable to provide a warning when an individual is in a particular area. Thus spatial coordinates in a chosen range may evoke some response, such as an alarm

Although the aforegoing description has used acoustic detectors operating at ultrasonic frequencies, in principle the system could work equally well with detectors

operating in other domains, e.g. rf. infra-red or even visible frequencies.

Claims

1. A method of tracking a mobile target within a monitored space using at least three fixed directional detectors having respective responsivities which vary

with angular position around the detector and being positioned around a reference axis with their respective detection axes angularly spaced apart and inclined towards

said monitored space such that the detectors have angularly overlapping response characteristics and a combined non-uniform responsivity profile;

the method comprising: emitting a signal at a source on the mobile target so that the at least three detectors provide respective electrical output signals (R, G, B); effecting chromatic processing to transform the output signals (R, G, B) into at least first and second chromatic parameters (H_s, L_s) whose values are dependent

upon the position of the target in said monitored space; and

2. A method as claimed in claim 1, which includes obtaining and establishing a store of values of angular position (θ) of the source relative to a

reference direction based upon values of the first chromatic parameter (H_s) and

obtaining and storing values of radial distance (r) of the source from said reference

axis based upon values of the second chromatic parameters (L,,). modified in

dependence upon values of the first chromatic parameter (H_s) to compensate for said

collective non-uniform responsivity profile of the detectors; and

comparing said stored values of angular position (θ) and radial distance (r) with values of the chromatic parameters (H, L)₅ calculated from measured values of

the detector output signals (R. G. B), to establish the instantaneous position of the

target.

3. A method as claimed in claim 1 or 2, wherein the signals (R, G, B) are

transformed into the chromatic parameters (H. L, S) from the equations :-

( 60(G-B) /(max(R,G,B)-min(R,G,B)), if max(R,G,B)=R H = \ 60(2+(B-R)) /(max(R,G,B)-min(R,G,B)), if max (R₅G₅B)=G f 60(4+(R-G)) /(max(R,G,B)-min(R,G,B)), if max (R₅G₅B)-B

R + G + B

maxfR. G. BVminfR. G. B)

S = max(R, G₅ B)+min(R, G, B)

4. A method as claimed in claim 1 or 2, wherein the signals (R₅ G, B) are transformed into the chromatic parameters (H, L₅ S) by the equations:- r=R-min(RGB) g=G-min(RGB)

b=B-min(RGB)

Lightness L = max(RGB) ÷ mϊn(RGB) _{v o} max(RGB) - mini RGB^') _^

Saturation S = _{/ D}^,_m , : _{/ Dr},^ ταsύi(RGB) ÷ -(Rm(RGB)

5. A method as claimed in any one of claims 1 to 4, which includes the

use of a fourth fixed, directional detector (M₄) having a responsivity which varies with angular position around the detector and which provides an output signal (F), the fourth detector being disposed with its detection axis inclined towards said

monitored space at an angle different to that of the first mentioned three detectors.

6. A method as claimed in claim 5 wherein the response characteristic of the fourth detector overlaps the response characteristics of all three of the first mentioned detectors.

7. A method as claimed in claim 5 or 6, wherein an additional chromatic parameter (S_F) is formed from the output signals (R, G, B, F) of the four detectors,

whose value is dependent upon the height of the target in said monitored space.

8. A method as claimed in claim 7 wherein the additional chromatic

parameter (S_F) is defined by:

S_F = greatest of(R, G, B) - F greatest of R, G_; B + F

9. A method as claimed in claim 7 or 8, wherein values of the fourth

chromatic parameter (S_F) are calibrated against measured values of the height of the

target in said monitored space and stored, for example in a look-up table.

10. A method as claimed in claim 7, 8 or 9, wherein the \ slues of tte

fourth chromatic parameter (S₅) are modified by a second calibration factor (C) dependent on the value of the first chromatic factor (Hs) to compensate for the

variability of the amplitude of the further chromatic parameter (S_F) with the value of

θ.

11. A method as claimed in any of claims 5 to 10, wherein the fourth detector (M₄) is disposed with its detection axis vertical.

12. A system for establishing the position of a mobile target within a monitored space, comprising:

at least three fixed directional detectors having respective responsivities

which vary with angular position around the detector and which are adapted to be positioned around a reference axis with their respective detection axes angularly spaced apart and inclined towards said monitored space such that the detectors have angularly overlapping response characteristics and a combined non-uniform responsivity profile; the at least three detectors providing respective electrical output signals (R₅

G₅ B); means for effecting chromatic processing to transform the output signals (R₅

G, B) into at least first and second chromatic parameters (Hs, Ls) whose values are

dependent upon the position of the target in said monitored space; and

means for determining the position of the target based on the values of the

chromatic parameters (Hs, Ls).

13. A system as claimed in claim 12. including means for obtaining values

of the first and second chromatic parameters (Hs₃ Ls) calibrated such that the value of the angular position (θ) of the source from a reference direction corresponds to the

value of the first chromatic parameter (Hs), and the value of the distance (r) of the

source from the reference axis corresponds to the value of the second chromatic

parameter (Ls), modified in dependence upon the value of the first chromatic

parameter (Hs) such as to compensate for said collective non-uniform responsivity profile of the detectors.

14. A systems as claimed in claim 12 or 13, further comprising a fourth fixed directional detector having a responsivity which varies with angular position around

the detector and which produces an output signal (F), the fourth detector being disposed with its detection axis inclined towards said monitored space at an angle different to that of the first mentioned detectors.

15. A system as claimed in claim 14 wherein the response characteristic of the fourth detector overlaps the response characteristics of all three of the first mentioned detectors.

16. A system as claimed in claim 14 or 15 comprising means forming an

additional chromatic parameter (Sf) from the output signals (B, G, B, F) of the four

detectors, whose value is dependent upon the height of the target in said monitored

space.

17. A system as claimed in any of claims 14 to 16. wherein the detection axes

of said at least three detectors lie at an acute angle to the vertical, preferably at

approximately 45° to the vertical.

18. A system as claimed in any of claims 14 to 17 wherein the detection axis

of the fourth detector is substantially vertical.

19^. A system for tracking a mobile target substantially as hereinbefore described with reference to and as illustrated I_n the accompanying drawings.

20. A method for tracking a mobile target substantially as here_inbefore described with reference to the accompanying drawings.