Method for determining the presence of a vehicle in a parking space

Abstract

A method for determining the presence of a vehicle in a parking space in a car park comprises: performing (E1) ultrasonic echo measurements, two successive measurements being separated by a predetermined time and forming an effective series of measurements; transmitting (E2) each series of measurements to a processing unit; comparing (E3) each measurement to the previous measurement within the same series; assigning (E4) each series of measurements to a group of similar series of measurements, a similarity score being calculated between the effective series and each group; if the last series of measurements performed is assigned to the group of similar measurements containing the most measurements, assigning (E5) an initial "available" status; and displaying (E6) the status assigned to the space. Abstract: Figure 2

Description

Title of the invention: Method for determining the presence of a vehicle in a parking space technical field

[0001] The present invention relates to the field of counting and guiding users in a parking lot.

[0002] More specifically, the invention relates to counting and guidance systems for users in a parking lot with a decentralized and autonomous architecture. PRIOR TECHNOLOGY

[0003] As is known, a conventional system for counting and guiding users in a parking lot consists of a network of equipment, including vehicle presence detection devices and display devices, connected within a CAN (Controller Area Network) communication network. Such equipment generally consists of ultrasonic transducers, configured to emit an ultrasonic wave and receive an echo of said wave reflected by an obstacle. Based on the spectral characteristics of the received echo, the counting system can then determine whether or not an obstacle, in this case a vehicle, is present in a parking space. The reception of an echo is then referred to as a "measurement."

[0004] In a parking lot environment, numerous factors can influence the spectral characteristics of an echo upon its reception by a detector: waves emitted by neighboring transducers or by other sound sources such as vehicles moving in the parking lot, variations in topography (irregularities in the ground, pillars, pipes, etc.) or variations in atmospheric conditions, etc. Naturally, these factors are compounded by the inherent defects of the equipment itself, which can cause the strength of an echo to vary by 10% to 30%.

[0005] Furthermore, when the target is a vehicle, the received echo must necessarily take into account the shape of said vehicle, which varies in particular from one model to another and from one type of vehicle to another. For example, the canvas roof of a convertible will absorb an ultrasonic wave emitted by the transducer, whereas a polycarbonate roof may reflect an ultrasonic wave towards the parking garage ceiling.

[0006] Thus, determining the availability of a parking space for each requires adjusting the transducer to avoid as many fixed obstacles as possible, noting and ignoring remaining fixed echoes, and taking into account local atmospheric conditions. Despite these precautions, which are already logistically and economically costly, the factors described The above cannot always be taken into account, thus affecting the quality of the measurements taken and the relevance of the counting system.

[0007] The invention aims to remedy all or part of the aforementioned drawbacks of the prior art, by proposing a method to substantially simplify the determination of the availability of a parking space, and therefore the reliability and cost of a counting system. PRESENTATION OF THE INVENTION

[0008] More specifically, the invention relates to a method for determining the presence of a vehicle in a parking space implemented by a counting and guidance system for users in a parking lot, the system comprising at least one presence detector opposite a parking space in said parking lot, the method comprising: • the performance of ultrasonic echo measurements via the presence detector, two successive measurements being separated by a predetermined duration and forming an effective series of measurements; • the transmission of each series of measurements to a processing unit for storage in memory; • the comparison of each measurement to the previous measurement within the same series; and, in the case where a series of successive measurements are not strictly identical, the procedure includes: • the assignment of each set of measurements to a group of similar sets of measurements, based on a plurality of similarity scores, a similarity score being calculated between the actual set and each group by the processing unit; • in the event that the last set of measures carried out is assigned to the group of similar measures comprising the most measures, the assignment of a first status "available" to said place and the assignment of a status "not available" to said place in all other cases; • displaying the status assigned to said space for the attention of parking users.

[0009] Thanks to such a combination of features, such a method makes it possible, with a high degree of reliability, to determine the status of a parking space, taking into account external factors that may influence the shape of the measured echo. Furthermore, such a method has the advantage of not requiring the storage of a large number of measurements in memory, thus accelerating the implementation of the method and improving the performance of the counting system, the benefit being directly proportional to the number of presence detectors installed.

[0010] Advantageously, a similarity score r for a data vector from a series of effective measures with respect to a data vector from a reference series of a similarity group is defined by [Math 1] YY f A^lmir / ) , with The average of all values ​​stored in the vector corresponding to the series of effective measurements A and / the average of all values ​​stored in the vector corresponding to the first series of measurements I.

[0011] Advantageously, a predetermined maximum number of measurements is stored in memory by the processing unit, with the oldest echo measurement being erased when the most recent measurement is transmitted to the processing unit. In such a configuration, the number of measurements stored in memory is further advantageously reduced, improving the performance of the counting system.

[0012] Advantageously, a second status is obtained by an inference made by the processing unit on the last echo measurement made, a final status "available" or "not available" of the place being obtained by unanimous decision of the first status and the second status.

[0013] According to another aspect of the invention, it relates to a system for counting and guiding users in a parking lot, comprising a plurality of vehicle presence detectors, each presence detector being opposite a parking space in said parking lot, the counting system being configured to implement the vehicle presence determination method described above for each parking space in said parking lot. PRESENTATION OF THE FIGURES

[0014] The invention will be better understood upon reading the following description, given solely by way of example, and referring to the accompanying drawings given by way of non-limiting examples, in which identical references are given to similar objects and on which:

[0015] Fig. 1 is a schematic representation of a counting and guidance system for users in a parking lot, according to a first aspect of the invention.

[0016] Fig. 2 is a flowchart describing the steps of a process for determining the presence of a vehicle in a parking space implemented by the counting system of Fig. 1.

[0017] It should be noted that the figures set out the invention in detail to enable implementation of the invention; although not limiting, said figures serve in particular to better define the invention where appropriate. DETAILED DESCRIPTION OF THE INVENTION

[0018] Figure 1 illustrates a system 1 for counting and guiding users in a parking lot according to a first aspect of the invention. The system 1 of Figure 1 comprises at least one first modular electronic card 2 connected to a power supply 5 having a first voltage. The power supply 5 is configured to provide the first voltage to the modular electronic card 2. Advantageously, if the power supply 5 provides a mains voltage of 230V AC, the system 1 includes a first transformer configured to provide the electronic cards 2 with the first voltage of 48V DC. The first transformer is located in the immediate vicinity of each modular electronic card 2, so as to form a "node" with said modular electronic card 2.

[0019] In [Fig. 1], each modular electronic card 2 is connected to at least one presence detector 9, each presence detector 9 being positioned opposite a parking space in the parking lot where the counting system 1 is installed. The presence detectors 9 are configured to detect the presence of a user's vehicle in a given parking space, so as to know in real time the occupancy capacity of the parking lot in which the system 1 is installed. In particular, the presence detectors 9 include ultrasonic transducers, configured to emit an ultrasonic wave and configured to measure an echo, in particular a reflected ultrasonic wave. Preferably, the transducers emit an ultrasonic wave at a frequency between 20 and 60 kHz. Following emission, the transducers measure an echo, in particular by applying a spectral filter to isolate the emitted frequency, here between 20 and 60 kHz.

[0020] The modular electronic boards 2 are configured to supply the second voltage to the presence detectors 9. The presence detectors, together with the modular electronic boards 2, form a CAN communication network. The CAN communication network is connected to a remote processing unit 13, i.e., a server.

[0021] According to the invention, the modular electronic boards 2 are connected to the presence detectors 9 via network cables 3. The network cables 3 are thus configured to transmit power and to connect the detectors 9 and the modular electronic boards 2 within the CAN communication network. In the illustrated embodiment, the network cables 3 are Category 5 Ethernet cables with a gauge of 24 AWG. The network cables 3 are plugged into one of the connectors of each modular electronic board 2. The network cable 3 advantageously has plugs comprising eight conductors; six of the eight conductors of said plug are dedicated to the transfer of electrical power, and two of the Eight conductors are dedicated to connecting the equipment within the CAN communication network. The microcontroller of the modular electronic boards 2 can thus communicate with the presence detectors 9 to ensure the operation of the counting system 1. Finally, each modular electronic board 2 can be connected by a network cable 3 to a dynamic display module 11, allowing parking lot users to view real-time information on the availability of parking spaces.

[0022] The CAN communication network has, in particular, characteristics falling within the scope of ISO 11898-3 (2006) and having an autonomous and decentralized architecture.

[0023] Advantageously, the system 1 may include a second modular electronic board 2 connected to the power supply in a similar manner to the first modular electronic board 2. Thus, the second modular electronic board 2 is configured to supply the second voltage to at least one group of presence detectors 9. This configuration is particularly advantageous for powering presence detectors 9 located furthest from the first modular electronic board 2, as the second electronic board 2 allows for the reinjection of a suitable electrical voltage. Furthermore, the second modular electronic board 2 also allows for re-amplification of the CAN communication network signal. It is then possible to compensate for line losses, particularly those due to Joule heating.

[0024] Figure 2 is a flowchart describing the steps of a method for determining the presence of a vehicle in a parking space implemented by the counting system illustrated in Figure 1, according to a second aspect of the invention. Preferably, the counting system is configured to implement the method for determining the presence of a vehicle for each parking space in said parking lot.

[0025] The method includes a first step El consisting of carrying out ultrasonic echo measurements via the presence detector 9, two successive measurements being separated by a predetermined duration and forming an effective series of measurements.

[0026] The process includes a second step E2 consisting of transmitting each measurement to a processing unit 13 for storage in memory.

[0027] The method includes a third step E3 consisting of comparing each echo measurement to the previous measurement. Such a comparison makes it possible, in particular, to determine whether the two curves are strictly identical, for example by calculating an autocorrelation coefficient according to a method known in the field of signal processing.

[0028] In the case where a series of successive measurements are not strictly identical, the method includes a fourth step E4 consisting of assigning each series of measurements to a group of similar series of measurements, based on a plurality of similarity scores r, the series of measurements being assigned to the group for which the series' similarity score r is the highest. The series of successive measurements preferably comprises more than two measurements, for example, five measurements. If a given echo measurement has the same similarity score r as several different groups of measurements, said measurement is simply assigned to the first group encountered by the processing unit 13. If a similarity score r, lower than a predetermined threshold, is determined for the series of measurements, a new similarity group is created and said measurement is assigned to said newly created group.

[0029] The remainder of the description focuses on the assignment step based on the calculated similarity scores. During a preliminary initialization phase, at least one initial set of measurements I is used to create a first group and stored as a data vector, in particular a 2-dimensional vector. Subsequently, an actual set of measurements A, stored as a data vector, in particular a 2-dimensional vector, is processed. The following example is given for a set of measurements and a similarity group, it being understood that the operation must be repeated so that it is possible to determine the highest similarity score for the actual set of measurements A with a plurality of groups.

[0030] The similarity calculations use the autocorrelation coefficient between two measurements from the newly acquired effective measurement series A and the first measurement series I, assigned to the first group. Let r be the similarity score between the effective measurement series A and the first measurement series I; then:

[0031] [Math 2] , f = ..................................................=■

[0032] with A the average of all the values ​​stored in the vector corresponding to the series of effective measurements A and / the average of all the values ​​stored in the vector corresponding to the first series of measurements I.

[0033] Advantageously, when a group comprises several sets of measurements, the sets within the same group are compared using their similarity score r, and the measurement with the highest similarity score r becomes a reference set for the group. The reference set is thus representative of the group and allows for the assignment of subsequently acquired sets to each group during the fourth step, E4. The reference set can be updated each time a set is added to a similarity group.

[0034] Advantageously, a training database is formed from each group, each series being stored in the database as a data pair comprising, for a given series of measurements, the data vector of said series of measurements and the data vector of the reference series of the group to which said series of measurements belongs. In such a configuration, a deep neural network of the processing unit 13 can be trained on the database thus formed, said neural network being able to be distributed across the microcontrollers of the electronic boards 2.

[0035] Advantageously, the deep neural network can be used to manage the groups and assign the series to each group itself, based on its own classification resulting from its supervised training.

[0036] In the case where the last set of measurements performed is assigned to the group of similar measurements comprising the most measurements, the method includes a fifth step E5 consisting of assigning a first status "available". In all other cases, that is to say when the last measurement is assigned to a group of similar measurements not comprising the most measurements, a first status "not available" is alternatively assigned to that place.

[0037] The method includes a sixth step E6 consisting of displaying the status assigned to the parking space to parking users. This display is achieved, in particular, by means of an indicator light showing a first color for the "available" status and a second color for the "unavailable" status. Finally, a counter in the dynamic display module 11, displayed to parking users, is incremented for each parking space with the "available" status and decremented for each parking space with the "unavailable" status. Thanks to this combination of features, this method makes it possible, with a high degree of reliability, to determine the status of a parking space, taking into account external factors that may influence the shape of the measured echo.Furthermore, such a process has the advantage of not requiring the storage of a large number of measurements in memory, accelerating the implementation of the process and improving the performance of the counting system 1, the benefit being directly proportional to the number of presence detectors 9 installed.

[0038] Advantageously, a predetermined maximum number of measurements is stored in memory by the processing unit 13, the oldest echo measurement being erased when the most recent measurement is transmitted to the processing unit 13. In such a configuration, the number of measurements stored in memory is further advantageously reduced, improving the performance and energy consumption of the counting system 1.

[0039] Advantageously, a second status is obtained through an inference performed by the deep neural network trained using the previously described database on the last echo measurement taken, a final "available" or "unavailable" status of the space being obtained by unanimous decision of the first and second statuses. In such a configuration, the reliability of the system is further enhanced, as the status of a given space is subject to two methods of determination.

[0040] It should also be noted that the invention is not limited to the embodiments described above. Indeed, it will be apparent to a person skilled in the art that various modifications can be made to the embodiments described above, in light of the information just disclosed to them.

[0041] In the detailed presentation of the invention given above, the terms used shall not be interpreted as limiting the invention to the embodiment set forth in this description, but shall be interpreted as including all equivalents which can be foreseen by a person skilled in the art by applying their general knowledge to the implementation of the teaching which has just been disclosed to them.

Claims

Demands

1. A method for determining the presence of a vehicle in a parking space implemented by a system (1) for counting and guiding users in a parking lot, the system (1) comprising at least one presence detector (9) opposite a parking space in said parking lot, the method comprising: • taking (E1) ultrasonic echo measurements via the presence detector (9), two successive measurements being separated by a predetermined time and forming an effective series of measurements; • transmitting (E2) each series of measurements to a processing unit (13) for storage in memory; • comparing (E3) each measurement to the previous measurement within the same series;and, in the case where a series of successive measurements are not strictly identical, the process includes: • the assignment (E4) of each series of measurements to a group of similar series of measurements, according to a plurality of similarity scores, a similarity score being calculated between the actual series and each group by the processing unit (13); • in the case where the last series of measurements carried out is assigned to the group of similar measurements comprising the most measurements, the assignment (E5) of a first status "available" to said space and the assignment of a status "not available" to said space in all other cases; • the display (E6) of the status assigned to said space for the attention of parking users.

2. A method for determining vehicle presence according to claim 1, wherein a similarity score r for a data vector of a series of effective measurements with respect to a data vector of a series of references from a similarity group is defined by Eh,E„(a™- , with 'a the mean of all the values ​​stored in the vector corresponding to the series of effective measurements A and / the average of all values ​​stored in the vector corresponding to the first series of measurements I.

3. A method for determining vehicle presence according to any one of the preceding claims, wherein a predetermined maximum number of measurements is stored in memory by the processing unit (13), the oldest echo measurement being erased when the most recent measurement is transmitted to the processing unit (13).

4. A method for determining vehicle presence according to any one of the preceding claims, wherein a second status is obtained by an inference made by the processing unit (13) on the last echo measurement made, a final status of "available" or "unavailable" of the space being obtained by unanimous decision of the first status and the second status.

5. System (1) for counting and guiding users in a parking lot, comprising a plurality of vehicle presence detectors (9), each presence detector (9) being opposite a parking space in said parking lot, the counting system (1) being configured to implement the method of determining vehicle presence according to any one of the preceding claims for each parking space in said parking lot.