FR3074383A1 - METHOD FOR RECEIVING A LIGHT SIGNAL MODULE VLC SIGNAL TYPE - Google Patents

Abstract

A method of receiving a VLC signal, comprising the steps of: - acquiring an image comprising fringes corresponding to high states and low states of the modulated light signal; converting the image into a signal of average intensities, the average intensity signal comprising components each representative of an average of luminous intensity of a row of the image; - assimilate the signal of average intensities to a time series, and implement a deseasonalization method on the time series, to obtain a curve recentered amplitude (26); - determining peak widths (27) of the amplitude recentering curve (26), by performing zero crossing detection of the amplitude recenter curve (27); reproducing a digital signal contained in the modulated light signal from the widths of the peaks (27).

Description

The invention relates to the field of methods for receiving modulated light signals of the VLC signal type (acronym for English Visible · bight Communication).

BACKGROUND OF THE INVENTION

The emission and reception of modulated light signals of the VLC signal type constitute a particularly promising technology for implementing an internal (or indcor, 1Q) or external (or büida & r) geolocation function of an electronic device.

Interior geolocation consists in providing the position of the electronic device (and therefore of its user) while it is in a room in any building: museum, station, airport, shop, workspace, etc.

The electronic device in question is for example a mobile phone, a tablet, a laptop, a connected watch, etc.

The interior geolocation · of the electronic device uses a lamp · with light-emitting diodes (here called LED lamp) positioned in the room to emit modulated light signals intended for the electronic device. The LED lamp thus performs both a room lighting function and the interior geolocation function.

The modulated light signals include an identifier of the LED lamp. The identifier contains • LED lamp position data. Position data can directly contain the position of the LED lamp. Alternatively, the position data can allow the electronic device to obtain the position of the LED lamp, which is stored for example in a server accessed by the electronic device. Of course, it is possible to geotag an electronic device using position data from a plurality of LED lamps.

Many technical challenges arise for the designers of such systems using the emission and reception of modulated light signals.

One of these challenges concerns the reception of modulated light signals. Most of the existing electronic devices mentioned earlier are not equipped with a device for receiving modulated light signals. One can therefore either provide the electronic device with additional reception equipment connected to a port of the device, for example to a female jack, or use a camera of the electronic device. In the latter case, the reception of the modulated light signals requires · to carry out · a processing of an image produced by the camera, which -must be particularly efficient so that · the reception of the modulated light signals is sufficiently effective.

OBJECT OF THE INVENTION

2) 5 The object of the invention is to improve the reception of modulated light signals of the VLC signal type.

SUMMARY OF THE INVENTION

With a view to achieving this goal, a method of receiving a modulated light signal of the VLC signal type is proposed, implemented in an electronic device comprising a camera comprising a CMO sensor), the method of reception including the steps of:

- acquire an image comprising fringes corresponding to high states and low states of the modulated light signal;

converting 1 / image into a signal of average intensities, the signal of average intensities comprising components each representative of an average of intensity: light · from a row of the image;

- assimilate the signal of · average · intensities to a time series, and implement a method of seasonal adjustment on the time series, to obtain a curve centered in amplitude;

- determining the widths of peaks of the curve centered in amplitude, by carrying out a detection of zero crossings of the curve centered in amplitude;

- reproduce a digital signal contained in the modulated light signal · from the widths of the peaks.

the reception method according to the invention amé20 significantly limits the reception of modulated light signals of VLC signal type.

An electronic device is also proposed, comprising a camera comprising a CMOS sensor and a processing component in which the reception method which has just been described is implemented.

The invention will be better understood in the light of the following description of a particular non-limiting embodiment of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

Reference will be made to the appended drawings, among which:

- Figure 1 shows an LED lamp and a user with a mobile phone;

- Figure 2 shows steps of an emission method;

- Figure 3 shows an identifier frame;

- Figure 4 shows an identifier frame comprising a first coded signal;

- Figure 5 shows a modulated identifier frame comprising a second coded signal;

- Figure 6 shows an image produced by a camera of the mobile phone ·;

- Figure 7 shows an object from Figure 6 and corresponding to an LED lamp;

- Figure 8 shows steps of a decoding phase of the method for receiving a modulated light signal of VLC signal type according to the invention;

- the figure; 9 represents a curve of a signal of average intensities as a function of the rows of an image portion;

- Figure 10 shows a curve centered in amplitude obtained from the curve of Figure 9;

- Figure 11 shows a Manchester decoding table;

FIG. 12 represents a curve of a signal of average intensities as a function of the rows of an image portion;

- Figure 13 shows a curve centered in amplitude, obtained from the curve of Figure 12 using a seasonal adjustment method using polynomial regression;

- Figure 14 represents a curve centered in amplitude, -obtained thanks to a- seasonal adjustment method using a conventional Hodrick-Prescott filter j

FIG. 15 represents a curve centered in amplitude, obtained by means of the reception method according to the invention;

- Figürë 16 represents a first curve, a second curve and a third curve illustrating an error rate of reception of the identifier as a function of a distance, the first curve corresponding to the use of a seasonal adjustment method using a polynomial regression, the second curve corresponding to the use of a seasonal adjustment method using a conventional Hodrick-Prescott filter, the third curve corresponding to the use of a seasonal adjustment method used in the reception method according to 1 ' invention.

DETAILED DESCRIPTION OF THE INVENTION

A method of transmitting a light-modulated signal of the VLC signal type is first described. The transmission method is implemented in a VLC signal transmission device.

The VLC signal transmission device comprises a light-emitting diode and a driving module intended to drive the light-emitting diode. The control module comprises a first processing component suitable for executing instructions for setting the steps of the transmission method, and component

piloting of the diode or dri ver, in English.

The first processing component here is a microcontroller, but could be a different component, for example a processor or an FPGA.

The driver here is a voltage driver,

With reference to FIG. 1, the VLC signal transmission device is integrated · into an LED lamp 1. The LED lamp 1 is here positioned on the ceiling of a room in any building: museum, station, airport , store, workspace, etc.

The LED lamp 1 uses the light emitting diode to light its environment. The LED lamp 1 thus fulfills a function of lighting the room.

The LED lamp 1 also fulfills an internal geolocation function of an electronic device provided with a device for receiving VLC signals. The electronic device is in this case · a mobile telephone 2 with which a user 3 is provided.

The interior geolocation function 20 operates in the following manner. The VLC signal emitting device of the LED lamp 1 continuously transmits emitted VLC signals comprising an identifier of the LED lamp 1. The identifier of the LED lamp 1 comprises position data of the LED lamp 1 25 in the room.

The emission of emitted VLC signals consists in switching the light-emitting diode ·, so as to switch a light power produced by the light-emitting diodeGente to generate binary modulated light signals constituting the emitted VLC signals.

• To switch the light-emitting diode, the control module selectively supplies the light-emitting diode with a supply current higher than a predefined high current threshold or a supply current. below a redefined low current threshold p.

The supply current greater than the predefined high current threshold makes it possible to generate a high state of a VLG signal transmitted, ie a binary value equal to "1", while the supply current lower than the predefined low current threshold allows to generate a low state of the transmitted VLC signal, ie a binary value equal to "0".

When the mobile telephone 2 receives the transmitted VLG signals, and therefore the identifier of the LED lamp 1, the mobile telephone 2 uses the position data of the LED lamp 1 to determine its own position in the room, and therefore the position of user 3 in the room.

With reference to FIG. 2, the transmission method first comprises the step of generating an original digital signal (step E1). The original digital signal first contains data to be transmitted, in this case the identifier of the LED lamp 1. The identifier is composed of a plurality of • bits forming one or more ^ bytes. The duration of a high state or a low state of the identifier is equal to T (period).

The original digital signal also contains a checksum (or checksuiri, in English). The transmission method thus comprises the step of calculating a checksum on the identifier (step Ë2). the calculation of the checksum) here consists in carrying out a binary addition of the bits ^; of the identifier taken two by two. A first intermediate sum is thus obtained composed of a plurality of bits. Then, each bit of the first intermediate sum is replaced by the opposite bit, to obtain a second intermediate sum. Finally, we take the last three bits of the second intermediate sum to obtain the checksum :.

An example is illustrated of obtaining the checksum.

The identifier ID of the LED lamp 1 is:

Id = Obi0101010.

The first intermediate sum Si is such that:

If 0bl0 + = + 0bl0 0bl0eûbl0 == 0bl000.

By taking the Opposite bits, we obtain the second intermediate sum S ₂ :

S ₂ = 0b0111.

The checksum Os is therefore:

Cs = lll.

The transmission method then comprises the step of coding the original digital signal, that is to say the identifier and the checksum, using a first coding, to obtain a first coded signal (step E3). The first coding is here a ster Channel coding. A bit equal to Q of the identifier: and of the checksum is coded by a transition from a low state to a high state, while a bit equal to 1 is coded by a transition from a high state to a low state.

We note that: Manchester coding is: a special case of OOK modulation (for On-OfT Këyïng). Any other coding resulting from an OOK modulation could be used.

This gives a coded identifier Id _Ma nch and a coded checksum Cs _Manch .

The IdManch coded identifier is:

I d _Ma nch = 100110 0110 011001.

The coded checksum Cs _Manch is: Gs _Ma hch = 101010.

In the coded identifier and in; the coded checksum, the maximum duration of a high state or a low state is equal to] T, and the maximum transition frequency Ftran between two transitions (from a high state to a low state or from a low state to a high state) is such that ·

F _tran = 2 / T.

The transmission method then comprises the step of constructing an identifier frame. The construction of; the identifier frame consists in concatenating from the start bits, the coded identifier and the Coded checksum] (step Έ4). The identifier frame is a concatenated signal resulting from the concatenation of the start bits and the first coded signal.

The start bits Bd are here such as: Bd = 01110010.

It is noted that the starting bits comprise a succession of three consecutive "1s", that is to say three high states; 3T / 2 consecutive. As such an estate is; impossible in the coded identifier and in the coded checksum, due to the use of Manchester coding, the three consecutive “1s” make it possible to distinguish the starting bits and therefore the start of the identifier frame.

The identifier frame is therefore the following nu10 meric signal transmitted:

011100101001100110011001101010.

The transmission method then comprises the step of starting the transmission of the identifier frame (step 5 E5).

At the time of transmission of the identifier frame, the transmission method detects the low states of the identifier frame and therefore, in particular, of the first coded signal (step E6).

The identifier frame is coded using a second coding. The second coding consists in modulating each low state of the identifier frame at an Emoa modulation frequency (step E7). The modulation frequency Fmaa is greater than the maximum transition frequency F _tan . The modulation frequency Fftod is here a multiple of the maximum transition frequency F _{t £ an} . The modulation frequency F _mo a is here such that:

Fmod ⁼ 2Ftran ⁼ 4/1.

With reference to FIG. 3, there is thus obtained, from the identifier frame and 'therefore from the first coded signal, a second coded signal, which is a modulated identifier frame' 5,

With reference to FIGS. 4 and 5, it can therefore be seen that a low state 6 of duration T of an identifier frame (not modulated) containing the first coded signal Sel, is replaced by four · transitions 8 of a low state to a high state of the associated modulated identifier frame 10 containing the second coded signal 5c ₂ , and that a low state 12 of duration · ï / 2 is replaced by two transitions 13 from a low state to a high state of the modulated identifier frame 10.

The modulation ceases when a transition from a low state to a high state is detected in the identifier frame and therefore, in particular, in the. first coded signal (step ES). Then, the detection of a low state 5 of the identifier frame (not modulated) resumes (step E6).

The transmission process continues until the complete transmission of the modulated identifier frame (step · E9j, then resumes at step ES). The modulated identifier frame 10 is thus retransmitted continuously.

To transmit the modulated identifier frame, it is. thus necessary to supply the light-emitting diode with a significant average electrical power, greater than the electrical power which would be supplied in the case where only conventional Manchester coding (or another conventional coding) is used.

In fact, the use of a conventional Manchester coding would make it possible to obtain coded signals 20 constituted half by high states and half by low states, and therefore VLC signals - transmitted formed half by high states. and low states. The use of a conventional Manchester coding therefore amounts to supplying the light-emitting diode with a supply current greater than the predefined high current threshold for a total duration equal to half of the duration necessary for transmitting the VLC signals. and supplying the light-emitting diode with a supply current below the pre-defined low current threshold for a total time equal to half the time necessary to transmit the VLC signals.

With a classic Manchester coding, there then the average electrical power supplied to the light-emitting diode is therefore: equal to 50% of a maximum electrical power corresponding to a continuous emission of high states.

Here, by modulating the low states of the frame of identifier 7, and therefore of the first coded signal Sel, we obtain a modulated identifier frame: 10, and therefore a second coded signal Sc ₂ , whose high states are more numerous. and have a more significant total duration. The duration during which the light-emitting diode is supplied by a supply current greater than the predefined high current threshold is greater, and the electrical power: average supplied to the light-emitting diode is therefore increased. The power: average electric supplied · to the light-emitting diode is here equal / to 75% of the maximum electric power ·.

The modulated identifier frame 10 is therefore associated with an average electrical power supplied to the light-emitting diode of 75% of the maximum electrical power. By increasing the modulation frequency of the ba.s states, the average electric power supplied to the light-emitting diode is increased, which can reach / 95% of the maximum electric power.

The second: Coding: _is therefore intended to increase the average electrical power supplied to the light-emitting diode to generate the VLC signals transmitted.

The increase in the average electric power makes it possible to increase the average light power produced by the light-emitting diode and therefore by the LED lamp 1. It improves: thus

The lighting in the room in which the LED lamp 1 is positioned.

It is noted that, since the modulation frequency and the frequency band of the VLC signals transmitted are known, a low-pass filter can be used in the VLC signal reception device · of the mobile telephone 2 to filter the · VLC signals received and thus recovering the unmodulated identifier frame 7, simply coded in Manchester.

The method for receiving a modulated light signal of the VLC signal type according to the invention will now be described. The reception method according to the invention is implemented here in the mobile telephone 2 and allows the latter to receive the VLC signals transmitted.

The mobile telephone 2 includes a camera comprising a CMOS sensor (for Compleiaentary Jfe / tad-Wide-SemïcOTdüctor). The mobile phone VLC signal receiving device 2 includes the mobile phone 2 camera, which is therefore used to receive the received VLC signals. In addition to the camera, the receiving device includes a receiving module connected to the camera. The reception module comprises a second processing component adapted to execute · instructions of a program · to implement · the different stages of the reception process. The second processing component here is a processor, but could be a different component, for example a microcontroller or an FPGA.

The rolling shutter effect (or “rolling shutter effect”) of the CMOS sensor is used here to receive the received VLC signals.

With reference to FIG. 6, the camera makes it possible to obtain images similar to image 20. The image 20 comprises white fringes 21 and black fringes · 22. Each white fringe 21 corresponds 5 to a high state of a received VLC signal, while each black fringe 22 corresponds to a low state of the received VLC signal. We can therefore find the identifier frame modulated from image · 20.

The method for receiving a modulated light signal 10 of VLC signal type firstly comprises a preliminary step during which the camera is initialized. Initialization involves adjusting camera parameters, including exposure time, ISO sensitivity and resolution.

The exposure time is the time that the CMOS sensor is exposed to light. A short exposure time produces relatively dark images because the amount of light at which the CMOS sensor is exposed is small. Long exposure time · produces: relatively bright images: 20 because the amount of light to which the CMOS sensor is exposed is large. To clearly distinguish the fringes: white 21 from the fringes 25 black 22, and therefore the high states of the low states of the VLC signals received, it is preferable: to choose a Short exposure time.

ISO sensitivity determines the sensitivity of the camera to light. High ISO 30 sensitivity increases the sensitivity of the camera. By choosing a high ISO sensitivity, the detection of changes in light intensity and therefore the contrast of the image 20 is improved. It is therefore possible to detect VLC signals emitted by increasing the distance between the LED lamp 1 and the mobile phone 2. The ISO sensitivity of a conventional camera is typically between 100 and 800, but some recent mobile phones 2 are equipped with a camera having an ISO sensitivity of 1600 or even 3200.

Finally, it is advantageous to choose a resolution close to or equal to 640 × 480 in order to decode the 'VLC signals received in a relatively short time.

Once the camera has been initialized, an image 20 is processed to decode the received VLC signals.

It is possible that several light emission sources, that is to say here several LED lamps 1, emit VLC signals transmitted simultaneously in the room. In this case, if the camera 20 has a sufficient angle of view or if the distance between the camera and each LED lamp 1 is relatively large, the image 20 may contain white fringes 21 and black fringes 22 representative of the VLC signals emitted produced by more than 25 LED lamps 1.

The reception method in this case detects all the LED lamps 1 present in the image: 20. Then, the reception method successively decodes the VLC signals received from each LED lamp 30 1, starting with the lamp LED 1 closest to the center of image 20 and ending with the LED lamp 1 furthest from the center of image 20.

The reception method thus comprises a step of thresholding the image 20. The thresholding step consists in keeping, in the image 20, only the pixels having a light intensity greater than 5 a predetermined light intensity threshold. A plurality of preliminary forms is thus obtained.

Then, morphological operations: are implemented · on the image 20. The morphological operations: include morphological opening operations 10 and morphological closing operations. One thus fills-spaces of reduced size and deprived of pixels inside the preliminary forms, and one eliminates- sets of pixels of reduced size: and isolated near the preliminary forms »

With reference to _: FIG. 7, we thus obtain, in image 20, a plurality of objects similar to object 23. Each object 23 of image: 20 corresponds to one: lamp: with LED 1.

Then, a labeling algorithm with compo2: 0 related health is applied to each object in the image.

A centroid of each object 23 is then determined. A Euclidean distance between the: centroid of each object 23 and: the center of image 20 is calculated,

Qn thus obtains an object table 23. In the object table 23, the objects 23 are classified as a function of their distance from the center of the image 20, according to an order of increasing distance.

Once all the LED lamps 1 have been identified and: all the corresponding objects 23 have been stored in the table, a first image portion, corresponding to the object 23 located in the first position in the table d objects 23, is first extracted from image 20.

The first image portion therefore corresponds to the LED lamp 1 closest to the center of the image 20.

The first image portion is defined so that the LED lamp 1 is located in the center of the first image portion.

A decoding phase is applied to the first image portion, to decode the received VLC signal produced by the LED lamp 1.

your decoding phase is carried out by a decoder programmed in the mobile phone reception device 2.

The decoding phase is then repeated for each object 23 of the object table, and therefore for each LED lamp 1. Each LED lamp 1 is located in the center of an associated image portion. One portion of an image is of course an image. If only one LED lamp 1 is in the room and only one object 23 is detected in the image 20, it is possible either to carry out the decoding phase on the entire image 20, or to extract an image portion of the image 20 centered on the object 23.

With reference to FIG. 8, the decoding phase Pha firstly comprises a conversion step (step E20). The image portion · is a two-dimensional signal. Conversion involves converting the portion of the image to a one-dimensional signal. The decoder works mainly with one-dimensional signals. For this, the average of the light intensities of the pixels of each row of the image portion is calculated.

© n note here that, for each image portion, it is possible to sample the columns forming the image portion - to reduce the number of columns on which the conversion is carried out. By carrying out a sampling consisting in selecting a column - all · the M columns, and by considering that the total number of columns of an image portion is equal to n, the processing is carried out on a number j of columns, j being equal to the part; integer of the n / M ratio. © n thus considerably reduces the size of the image portion, and therefore the computation times are significantly reduced.

This gives a signal of average intensities, representative of the received VLC signal, comprising components each equal to an average of the light intensities of the rows of the image portion.

With reference to FIG. 9, the curve 24 of the signal of average intensities is formed by a succession of peaks 25. The peaks 25 include peaks of maximum local intensity and peaks of minimum local intensity. Each peak of maximum local intensity corresponds to a high state of the received VLC signal, and each peak of minimum local intensity corresponds to a low state of the received VLC signal. To obtain the number of bits corresponding to a high state or a low state, it is necessary to evaluate a width of each peak 25.

However, it can be seen in FIG. 9 that the amplitude of curve 24 has not passed over a value of constant amplitude. This is due to the fact that, as the LED lamp 1 is located in the center of there by image, the light intensity is greater at the center of the image portion than at the ends of: the port- image ion.

Thus, prior to the evaluation of the width of the peaks 25, the amplitude of the curve 24 of the signal of average intensities is centered: on a virtual zero value, to: obtain a curve centered in amplitude 26: d a signal centered in amplitude, visible in FIG. 10. The amplitudes of the peaks 27 are thus defined with respect to the same light intensity reference, corresponding to the virtual zero value 28, and the width of each peak corresponds to the width of the peak 27 at the level of this virtual zero value - 28.

-For this, we assimilate the: signal of average intensities to a time series, and we use: A method of seasonal adjustment of time series. A trend of the time series is then determined, then this trend is subtracted from the signal of average intensities to obtain a signal d: - 'average intensities re-centered.

The time series seasonal adjustment method uses a Hodrick-Presoott filter here.

It is therefore assumed that the signal of average intensities is a time series made up of · of value yt, and that each- value yt is equal to the sum of a component of tendency g _t and a cyclical component and, so that :

yt = gt + c _t , with t = 1, ..., T.

The determination: of cyclic components - c _t makes it possible to: produce the centered curve in amplitude 2 6.

The use of the Hodrick-Prescott filter has objective lice to calculate the cyclic components c _t from the values y _t by solving the minimization problem following r ^{+ λ} Σί = ιΚ5ί - 9t-i) ~ (5t-i “0t - ₂ ) l ² } ·

The minimization problem uses a smoothing parameter N.

The smoothing parameter · λ “penalizes” the variability of the trend component g _t . The higher the value of the smoothing parameter 1, the lower the variability of the trend component g _t , and the more the trend component is smoothed. On the contrary, the lower the value of the smoothing parameter λ, the greater the variability of the trend component g _t . When the value of the smoothing parameter 1 is close to zero, the component, of tendency g _t becomes equivalent to the value y _t .

The minimization problem which has just been mentioned can also be written as follows · y _T = (AF + l _T ) g _T , where γ _τ is a vector whose components are jg, gy is; a vector whose components are gt, ly is the identity matrix and F is the following pentadiogonal matrix:

Services -2 1 0 - ... ... o- -2 5 -4 1 0 ... ... ... ... ... 0 1 -4 6-4 1 0 ... ". ·. .. ·. 0 0 1 -4 g _4 1 0 ... 0 0 0 1 -4 6 _4 1 0 ...... 0 E = : : 0 ... ...... 0 1 -4 6-4 1 0 0 ... ... ... ... 0 1 -4 g _4 1 0 ... ...... ·· ... 0 1-4 5 -2 -0 ... ··· · ** ... 0 1-2 1-

So we have :

g _T = (IF + I _T ) ¹ y _T ,

And therefore :: · c _T = y _T - g _T , where c _T is is a vector whose components are les- c _t .

The value of the smoothing parameter K depends on the periodicity of the data: yt.

Here, as the LED lamp 1 is located at the center of the image portion, the intensity of the signal of medium intensities is relatively strong in the middle of the time series, and relatively weak: at the beginning and at the end of the time series. These differences in intensity tend to affect the efficiency of the Hodrick-Prescott filter.

To solve this problem, different: · smoothing parameter values are · used.

Considering that the time series y _t has a size equal to n, the smoothing parameter is the following smoothing vector:

1 2  800 800 not - + 1 8 1600 n ₇ + l 4 14400 n 5/3 ^{+ 1} 1600 n 4/3 ^{+ 1} 1100 not +1 20/17 500 not - 500 -

The smoothing parameter is thus a smoothing vector comprising a plurality of components.

The plurality of components includes central components, located at the center of the smoothing vector, and extreme components, located at the ends of the smoothing vector.

The central components have values greater than those of the extreme components.

A value of a central component is greater than 10000, and a value of an extreme component is less than 1000 ·

The values of the extreme components are here equal to 800 and 500. The values of the central components are here equal to 1100, 1600 and 14400.

Thus, the smoothing is relatively weak at the start and at the end of the time series, because the low light intensity of the received VLC signal results in a low amplitude of the signal of medium intensities, which requires a weak smoothing. On the contrary, the smoothing is relatively strong in the middle of the time series ^ because the high light intensity

23:

of the received VLC signal results in a large amplitude of the medium intensity signal, which requires strong smoothing.

This gives the refocused curve * in amplitude 26 of the signal refocused in amplitude (step E21 in FIG. 8).

It is then necessary to determine the * bits constituting the VLC signal received from the curve * centered in amplitude 26 and, in particular, the number of bits of each * sequence of high states and low states. The width of each peak 27 corresponds to

a certain number of bits equal to 1 (for a peak: of intensity local maximum ' ) or 0 (for a peak of intensity local Minimum *) *. 15 We measure for that the width of each peak

of the curve centered in amplitude 26 by performing a *: detection of zero crossings (or zero-crossing detection, in English) of the curve centered in amplitude: 26 (step E22).

Each peak of maximum local intensity corresponds, according to its width, to one of the following * sequences:

1, 11, 111.

Each cortes25 minimum local intensity peak, depending on its width, is one of the following sequences:

0.00.

A digital signal received representative of the received VLC signal is thus obtained.

Then, once * the sequences of high states and low states * have been decoded, the reception method * comprises the step of detecting the * starting bits (step E23). As a reminder, the start bits contain the sequence:

01110010.

A sliding window is applied to the digital signal received to detect the start bits.

Then; once the start bits have been identified, the bits of the coded identifier and the coded checksum can be recovered.

the bits of the digital signal received, which follow the start bits, undergo for this a Manchester decoding (step · E24). For this, 22 bits' - are here taken into account 'from the end of the' start bits.

Manchester 'decoding' is performed using the table in · Figure 11. If no error is identified during Manchester decoding, an eleven bit sequence is obtained in which the first eight bits correspond to the identifier (not coded) of the LED lamp 1 and the last three bits correspond to the checksum · (not coded ').

Manchester decoding makes it possible to detect an error if the digital signal received contains sequences of two consecutive "1" s or two consecutive "0" (except of course the start bits), since the decoding corresponding to the table is produced for pairs of two bits. However, even if no error is detected by Manchester decoding, it is possible that the digital signal received • contains an error.

The following Manchester encoded digital signal 30 may for example be transmitted:

01-01-10-01-10-10-10-01, which corresponds to the following uncoded digital signal:

00101110.

If the received coded digital signal is:

10-01-10-01-10-10-10-01, which corresponds to the following uncoded digital signal received:

1010) 1110, no error will be detected at the time of Manchester decoding, whereas · the received uncoded digital signal is different from the emitted non-coded digital signal.

The use of checksum avoids this type of error.

The calculation of the checksum again consists in carrying out a binary addition of the bits of the identifier taken two by two. A first intermediate sum is thus obtained composed of a plurality of bits. Then, each bit of the first intermediate sum is replaced by the opposite bit ·, to obtain a second intermediate sum. Finally, we take the last three bits of the second intermediate sum to obtain the checksum. It is checked whether the checksum obtained in the digital signal received corresponds to that of the digital signal transmitted (step 25 E25).

If the checksum corresponds, the reception method includes the step of providing the second processing component of the mobile telephone 2 with the identifier of the LED lamp 1, so that the geolocation of the mobile telephone 2 can be carried out.

Otherwise, an error message is generated by the second processing component (step E26).

Of course, all the steps of the decoding phase Ph_d are then carried out for each object 23 of the object table 23 (and therefore for each LED lamp 1), in an order corresponding to the position 5 of the objects · 23 in the object table 23.

Figures ^ 12 to 16 illustrate the performance of the reception process.

FIG. 12 represents a curve 30 of a signal of average intensities. FIG. 13 represents a curve centered in amplitude 31 obtained using a seasonal adjustment method using a polynomial regression. FIG. 14 represents: a curve centered in amplitude 32 obtained thanks to a seasonal adjustment method using a conventional HOdrick-Prescott filter. FIG. 15 represents a curve centered in amplitude 33 obtained thanks to the reception method described here.

Note that the curve centered in amplitude 31 and the curve centered in amplitude 32; are not centered on the virtual null value 34 between: the rows / 280 'to; 350, cohtrairemêht to the curve centered in amplitude 33, which is perfect alternating centered on the value 'null: virtual 35.

Figure 16 shows a first curve

36, one: second curve / 37 and a third curve 38.

The first curve 36, the second curve 37 and the third curve 38 illustrate a rate: error of reception of the identifier of the LED lamp 1 by: the telephone: 'mobile 2, as a function of the distance 30 between the mobile phone 2 and lamp / LED 1.

The first curve 36 corresponds to *

The use of a seasonal adjustment method

reading a polynomial regression. The second curve 37 corresponds to the use of a seasonal adjustment method using a conventional Hodrick-Prescott filter. The third curve 38 corresponds to the use of a seasonal adjustment method used in the reception method described here.

The average error rate is equal to 0.0093% for the curve · 36, to 0.0107% for the curve 37, and to 0.0030% for the curve 38. The reception process which has just been described is therefore very efficient, especially when the distance between the LED lamp 1 and the mobile phone 2 is large.

The invention is not limited to the particular embodiment which has just been described, but, on the contrary, covers any variant coming within the scope of the invention as defined by the claims.

Qn has described here that the second coding, intended to code the first coded signal with the aim of increasing an average electric power supplied to the light-emitting diode, consists in carrying out a frequency modulation on the low states of the first coded signal. This type of second coding is particularly advantageous in the case where the driver of the diode is a voltage driver. This second coding is also particularly advantageous in the case where a variator is positioned at the input of the driver, whether the driver is a voltage driver or a current driver.

In the case where the driver is a current driver.

and Where a variator is not positioned at the input of the driver, it is particularly advantageous to choose another type of second coding. The second coding then consists in modifying a duty cycle of the first coded signal (obtained following a Manchester · 'coding). We will increase, for example, a duration of the 5 high states of the first coded signal. This increases the duration of the current pulses which drive the diode ·, and therefore increases the average electric power · supplied to the light emitting diode and therefore the average light power produced · by the light emitting diode.

Although we have described here the implementation of the invention in an indoor geolocation function, the invention can of course be implemented in a different application, for example in an outdoor geolocation function /.

Claims (6)

1. Method of receiving a light signal 5 VLC signal type modulated ·, implemented in an electronic device (2) comprising a camera comprising a CMOS sensor, the reception method comprising the steps of: - acquire an image · (20) including fringes 10 (21, 22) corresponding to high and low states of the modulated light signal; converting the image into a signal of average intensities, the signal of average intensities comprising components each representative of an average Light intensity of a row of the image; - assimilate the signal of average intensities to a time series, and implement a · method of seasonal adjustment on the time series, to obtain a curve centered in amplitude (2 6); 20 - determining the widths of peaks (27) of the curve centered in amplitude (26) _z ën carrying out a detection of passages through zero of the curve · centered in amplitude (27); - reproduce a digital signal contained in the 25 light signal modulated from the widths of the peaks (27). 2. Reception method according to claim 1, in which the seasonal adjustment method implements a Hodrick ^ Prescott filter. 30 3). Reception method according to claim 2, in which the implementation of the HodrickPrescott filter uses a smoothing parameter, the smoothing parameter being a smoothing vector comprising a plurality of components. 4. Reception method according to claim 3, the plurality of components comprising components 5 central health and extreme components, the central components having values greater than those of the · extreme components. 5. 'Reception method according to claim 4, in which a value of a central component 10-is; greater than 10,000, and in which a value of an extremal component is less than 1,000. 6. 'reception method according to claim L, intended · to receive modulated light signals · generated' by separate light emission sources, The reception method further comprising a step of implementing · morphological operations on the image (20) to distinguish objects (23) each corresponding to a source; light emission. 7. Reception method according to claim 20 6, further comprising a step of applying a labeling algorithm for connected components on each object; (23). 8. Reception method according to claim 6, further comprising the steps of determining a 25 center of each object; (23), and to calculate a Euclidean distance; between the king center of each object (23) and the center of the image (20). 9. The reception method according to claim G, further comprising the step of extracting from the image (20) · image portions each comprising an object (23), each object (23) being positioned in the center ^ of the corresponding image portion. 10. Reception method according to one of the preceding claims, the modulated light signal comprising position data used to geolocate the electronic device. 11. Electronic device · comprising a camera comprising a CMOS sensor, and a processing component in which the reception method according to one of the preceding claims is implemented. 12. Electronic device according to claim 10 11, the electronic device being a mobile phone (2) or a tablet or a laptop or a connected watch.