Car washer

Abstract

To efficiently generate a vehicle type discrimination model which can be utilized by a car washing machine.SOLUTION: A model generation method includes the steps of: acquiring vehicle type information obtained from images of a vehicle (X) photographed by a camera (96); acquiring vehicle shape data generated based on output data of a sensor (vehicle shape detection sensor 52) moving relative to a vehicle in a longitudinal direction of the vehicle; and generating a vehicle type discrimination model (discrimination model M) in which the acquired vehicle shape data is an explanatory variable, and the acquired vehicle type information is an objective variable.SELECTED DRAWING: Figure 7

Description

Technical Field

[0001] The present disclosure relates to a method for generating a model used by a car washer and a car washer.

Background Art

[0002] In a car washer, vehicle washing is performed according to selections by a user's operation and detection results of sensors for detecting vehicle types provided in the car washer.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, among commercially available vehicle types, there are vehicle types that are not suitable for general course washing selected in a car washer. Therefore, it is desirable for a car washer to have a function of determining vehicle types. One aspect of the present disclosure aims to realize a method for efficiently generating a vehicle type discrimination model (vehicle type discrimination algorithm) that can be used by a car washer.

Means for Solving the Problems

[0005] A method for generating a model according to an aspect of the present disclosure includes: obtaining vehicle type information indicating a vehicle type of a vehicle to be washed by a car wash body that washes the vehicle while relatively moving with respect to the vehicle in the longitudinal direction of the vehicle, where at least a part of the vehicle type information is obtained from an image of the vehicle captured by a camera that captures the vehicle; obtaining vehicle shape data of the vehicle generated based on output data of a sensor for detecting the vehicle to be washed, the sensor having a plurality of optical axis sensors arranged parallel to each other along an optical axis and detecting the presence or absence of an object on the optical axis, and relatively moving with respect to the vehicle in the longitudinal direction of the vehicle; and generating a vehicle type discrimination model using the obtained vehicle shape data as an explanatory variable and the obtained vehicle type information as an objective variable.

Effects of the Invention

[0006] According to an aspect of the present disclosure, a vehicle type discrimination model that can be used by a car wash can be efficiently generated.

Brief Description of the Drawings

[0007]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Best Mode for Carrying Out the Invention

[0008] Among commercially available vehicle models, there are vehicle models equipped with parts that may be damaged in the brush control during general course washing selected in a car wash. Vehicle models that are not suitable for general course washing by a car wash are referred to as specific vehicle models. Conventionally, for such specific vehicle models, measures have been taken to prohibit the use of the car wash through notification on an explanatory signboard. However, if the use of the car wash is prohibited, the number of vehicles to be washed will decrease.

[0009] In addition, in order to enable the use of the car wash for specific vehicle models, there may be cases where measures are taken to avoid brushes at the parts where parts that may be damaged are installed in all vehicles to be washed. However, since brush avoidance is performed in all vehicles to be washed, the cleaning power may decrease and customer satisfaction may decrease.

[0010] Hereinafter, the configuration of a car wash that can perform washing by changing a part of general course washing at a specific part of a specific vehicle model, thereby reducing the possibility of parts being damaged in the specific vehicle model, will be described.

[0011] 〔Embodiment〕 The car wash and monitoring device according to one aspect of the present disclosure will be described. The car wash and monitoring device according to one aspect of the present disclosure are installed as a set at a gas station, an automobile repair shop, or the like.

[0012] In this specification and the drawings, the xyz coordinates are defined such that the xz plane is the horizontal plane. The positive direction of the x-axis is referred to as the front direction, and the negative direction of the x-axis is referred to as the rear direction. This is because in the car wash main body whose configuration will be described later, the side where the vehicle enters is the front and the side where the vehicle exits is the rear. Also, the positive direction of the z-axis is referred to as the right direction, and the negative direction of the z-axis is referred to as the left direction. For the y-axis parallel to the vertical direction, the positive direction is the upward direction, and the negative direction of the y-axis is the downward direction.

[0013] In the drawings, for a configuration that is vertically symmetric or horizontally symmetric with respect to a certain axis, in some cases, only one of the symmetric configurations is labeled with a reference sign, and the other symmetric configuration is omitted from being labeled with a reference sign.

[0014] <Car washer 2: Overall configuration> FIG. 1 is a schematic diagram showing a schematic side view 2S of a car washer main body 4 and a remote panel 6 included in the car washer 2 according to the present embodiment, and a schematic front view 4F of the car washer main body 4. FIG. 2 is a block diagram showing the configuration of the car washer 2 and the monitoring device 9 according to the present embodiment.

[0015] As shown in FIGS. 1 and 2, the car washer 2 according to the present embodiment includes a car washer main body 4 that washes a vehicle X, which is a vehicle to be washed. The car washer 2 further includes a remote panel 6 that acquires the car washing conditions of the vehicle X by the car washer main body 4. In the schematic side view 2S, the outer shape of the vehicle X is shown by a dotted line in order to show that the vehicle X is located deeper in the paper surface than the remote panel 6.

[0016] <Car washer 2: Car washer main body 4> As shown in the schematic front view 4F, the car washer main body 4 includes, for example, two frames 8 and a ceiling portion 10 that connects the upper ends of the two frames 8. The car washer main body 4 has a structure in which the vehicle X can pass through a space 4S surrounded by the frame 8 and the ceiling portion 10 along the entry direction DA of the vehicle X shown in the schematic side view 2S. In the present specification, the entry direction DA is defined as the direction from the front surface 4A to the rear surface 4B of the car washer main body 4.

[0017] The car washer main body 4 includes a drive unit 4X and a cleaning unit 4Y. The drive unit 4X includes wheels 12 provided at the lower parts of the respective frames 8 and a drive device 11 that rotationally drives the wheels 12. When the drive device 11 rotationally drives the wheels 12, the car washer main body 4 relatively moves in the front-rear direction with respect to the vehicle X along a rail R disposed on the ground G. The rail R is formed, for example, along the entry direction DA. Here, while relatively moving with respect to the vehicle X, the car washer main body 4 performs cleaning (car washing) on the vehicle X in the space 4S.

[0018] The cleaning unit 4Y includes a plurality of rotating brushes that slide on the vehicle X and perform brushing. For example, the cleaning unit 4Y includes a top brush 14, a side brush 16, and a rocker brush 18, each of which is rotated by a rotation motor (not shown). The top brush 14 slides along the upper surface of the vehicle X and cleans the upper surface of the vehicle X. The side brush 16 and the rocker brush 18 clean both side surfaces of the vehicle X.

[0019] The side brush 16 according to the present embodiment includes, for example, a rotation shaft 16S that extends in a direction substantially perpendicular to the ground G, and a plurality of bristles 16T formed on the rotation shaft 16S. In other words, the side brush 16 includes a rotation shaft 16S that is parallel to at least a part of the side surface of the vehicle X, and the bristles 16T that rotate around the rotation shaft 16S are brought into contact with at least a part of the side surface of the vehicle X to clean the side surface. Further, the side brush 16 may include an arm 16A that supports the rotation shaft 16S and moves the rotation shaft 16S in the left-right direction of the vehicle X by a power unit (not shown).

[0020] A tank storage portion 20 for storing a plurality of liquid storage tanks (not shown) storing various liquid agents including a detergent or a wax is arranged on a side portion of the car wash machine main body 4. Above the tank storage portion 20, a distribution pipe portion 22 for distributing water containing city water or a liquid agent from each liquid storage tank is provided. From the distribution pipe portion 22, a first clean water nozzle 24, a second clean water nozzle 26, a first detergent nozzle 28, a second detergent nozzle 30, a water repellent coating nozzle 32, and a wax nozzle 34 included in the cleaning unit 4Y are respectively led out via solenoid valves (not shown).

[0021] The first water purification nozzle 24 and the second water purification nozzle 26 are respectively arranged on the front surface 4A side and the rear surface 4B side of each frame 8 of the car washing machine main body 4, and spray water including city water onto the vehicle X. The first detergent nozzle 28 and the second detergent nozzle 30 are respectively arranged on the front surface 4A side and the rear surface 4B side of each frame 8, and spray a cleaning liquid including shampoo or the like onto the vehicle X. The water repellent coating nozzle 32 and the wax nozzle 34 are arranged on the rear surface 4B of the car washing machine main body 4. The water repellent coating nozzle 32 sprays a liquid agent of a water repellent coating agent onto the vehicle X. The wax nozzle 34 sprays wax onto the vehicle X.

[0022] In addition, a blower 36 for generating an air flow to dry the vehicle X is provided in the car washing machine main body 4. A top air blowing nozzle 38 and a side air blowing nozzle 40 are connected to the blower 36. The top air blowing nozzle 38 is provided at the upper center of the car washing machine main body 4 and blows air toward the ceiling surface of the vehicle X. The side air blowing nozzle 40 is provided on both sides of the car washing machine main body 4 and blows air toward the side surface of the vehicle X. The car washing machine main body 4 dries the washed vehicle X by the air blowing of the top air blowing nozzle 38 and the side air blowing nozzle 40.

[0023] An operation panel 42 is arranged on the front surface of one frame 8 of the car washing machine main body 4. The operation panel 42 is provided with operation buttons (not shown) for setting car washing conditions. For example, a user who has gotten out of the vehicle X or other technicians may operate the operation buttons to set car washing conditions or the like.

[0024] <Car washing machine 2: Control unit 44, Memory unit 45, Input / output unit 47> Furthermore, the car washing machine 2 is provided with a control unit 44 for controlling each part of the car washing machine main body 4. In the present embodiment, the control unit 44 includes a car washing control unit 441, an outer shape creating unit 442, a vehicle type discriminating unit 443, and a counting unit 444. The outer shape creating unit 442, the vehicle type discriminating unit 443, and the counting unit 444 will be described later. The car washing control unit 441 controls the washing of the vehicle X by the car washing machine main body 4 by controlling the operations of each part of the driving unit 4X and the cleaning unit 4Y of the car washing machine main body 4.

[0025] The car washer 2 also includes a storage unit 45 that stores various types of information, and an input / output unit 47 for inputting and outputting various types of information to and from external devices. The input / output unit 47 includes a communication interface for communicating with external devices either through a network or directly.

[0026] The control unit 44, the storage unit 45, and the input / output unit 47 may be provided in the car washer main body 4, or may be located outside the car washer main body 4. The control unit 44 may transmit and receive information between the car washer main body 4 or a remote panel 6 described later via the input / output unit 47, and control the car washer main body 4. The control unit 44 is composed of a processor such as a CPU, for example, and each control is realized by executing a control program stored in a memory on the processor.

[0027] <Car washer 2: Remote panel 6> The remote panel 6 is located, for example, on the front side of the car washer main body 4 and is arranged approximately along the moving direction of the car washer main body 4. Further, as shown in FIG. 1, the front surface of the remote panel 6 is arranged to face the side surface of the vehicle X before being washed by the car washer main body 4, in other words, before entering the interior of the car washer main body 4. For this reason, in FIG. 1, the back surface of the remote panel 6 is illustrated.

[0028] As shown in FIG. 1, the remote panel 6 includes a housing 46 and a support column 48 that stands on the ground G and supports the housing 46. The remote panel 6 may acquire at least a part of the car washing conditions of the vehicle X by the car washer main body 4 through operations such as a touch panel or buttons (not shown) provided on the housing 46. The control unit 44 may control the car washer main body 4 based on at least a part of the car washing conditions acquired by the remote panel 6 and wash the vehicle X.

[0029] <Car washer 2: Vehicle shape detection sensor 52> The car washer 2 further includes a vehicle shape detection sensor 52 (sensor) located in front of the car washer main body 4 for reading the vehicle shape (silhouette) from the side of the vehicle X as the car washer main body 4 moves. The car washer 2 also includes an encoder 13 connected to a drive device 11 that drives the wheels 12 of the drive unit 4X.

[0030] The vehicle shape detection sensor 52 may be a so-called area sensor including a plurality of photoelectric switches (optical axis sensors) each having a horizontal optical axis that crosses the relative movement path between the vehicle X and the car washer main body 4 at a right angle. By using such an area sensor, vehicle shape data of the vehicle X can be created.

[0031] The vehicle shape detection sensor 52 can obtain vehicle height data indicating the vehicle height of the vehicle X at the position of the vehicle shape detection sensor 52. As shown in FIG. 1, the vehicle shape detection sensor 52 is arranged opposite to the left and right side positions near the entrance of the car washer main body 4. As the car washer main body 4 moves relative to the vehicle X in the front-rear direction, the vehicle shape detection sensor 52 moves relative to the vehicle X in the front-rear direction and obtains detection signals from each photoelectric switch.

[0032] The detection signals from each photoelectric switch are an example of information regarding the outer shape of the vehicle X. The plurality of photoelectric switches are arranged, for example, at a predetermined interval (e.g., 15 mm) in the vertical direction, starting from a height of 650 mm from the ground G. The number of photoelectric switches is not particularly limited and may be, for example, 106. As the vehicle X moves relative to the car washer 2 in the front-rear direction, the vehicle height of the vehicle X at the position of the vehicle shape detection sensor 52 is detected by detecting the photoelectric switches whose optical axes are blocked by the vehicle X and the photoelectric switches whose optical axis blocking is released.

[0033] <Regarding the execution of the car washing operation> The control unit 44 of the car washer 2 controls the drive unit 4X and the cleaning unit 4Y based on at least a part of the car washing conditions acquired by the remote panel 6, the detection signals from each photoelectric switch of the vehicle shape detection sensor 52, and the pulse signal of the encoder 13, thereby performing car washing according to a predetermined procedure.

[0034] Specifically, the control unit 44 receives data necessary for car washing, such as setting signals for the selected car washing course pattern and operation signals such as start and stop, from the remote panel 6. Further, the control unit 44 receives detection signals of each photoelectric switch of the vehicle shape detection sensor 52 and pulse signals of the encoder 13. The car washing control unit 441 of the control unit 44 controls the drive unit 4X, the cleaning unit 4Y, etc. based on these signals from the remote panel 6 and the detection signals of the sensors, controls the operation of the entire car washing machine main body 4, and executes the car washing operation.

[0035] The control unit 44 includes an outer shape creation unit 442. The outer shape creation unit 442 acquires information regarding the outer shape of the vehicle X from each part of the sensor system 7 and creates outer shape data of the vehicle X. In the present embodiment, the sensor system 7 is composed of the vehicle shape detection sensor 52 and the distance sensor 71.

[0036] The distance sensor 71 is composed of an encoder 13 provided in the drive unit 4X of the car washing machine main body 4 and a count unit 444 provided in the control unit 44. The count unit 444 has a function of counting the pulse signals of the encoder 13. The count by the count unit 444 is reset when the vehicle X stops at a predetermined stop position and the car washing machine main body 4 starts to move toward the vehicle X.

[0037] For example, the car washing control unit 441 transmits a drive start instruction to start driving the wheels 12 to the drive device 11, and transmits a signal indicating that the car washing machine main body 4 has started to move to the count unit 444. When receiving the signal, the count unit 444 may reset the count. Based on the count by the count unit 444, the traveling distance of the car washing machine main body 4 is detected. The traveling distance of the car washing machine main body 4 is, in other words, the relative movement distance between the vehicle X and the car washing machine main body 4. The distance sensor 71 can output information indicating the traveling distance (relative movement distance) as traveling count data.

[0038] The outer shape creation unit 442 creates the outer shape data of the vehicle X from the detection signals of the photoelectric switches output by the vehicle shape detection sensor 52 and the travel count data output by the distance sensor 71. FIG. 3 is an example of an output result in which the outer shape data of the vehicle X created by the outer shape creation unit 442 is output as a graph. The outer shape data created by the outer shape creation unit 442 is, for example, data showing the silhouette when the vehicle X is viewed from the side as shown in FIG. 3.

[0039] The outer shape creation unit 442 can determine the tip position of the vehicle X as the position where the photoelectric switch is first shaded after the car washer main body 4 starts to move. For example, in FIG. 3, it can be seen that the optical axis is first shaded at the count (travel count) 15 counted by the counting unit 444. In this case, the outer shape creation unit 442 determines the position of count 15 as the tip position of the vehicle X.

[0040] When the outer shape creation unit 442 creates the outer shape data, it attaches the creation time and stores it in the storage unit 45 as the outer shape data log 453. All or required parts of the outer shape data log 453 stored in the storage unit 45 can be transferred to an external device through the input / output unit 47 as appropriate.

[0041] The vehicle type discrimination unit 443 discriminates the vehicle type of the vehicle X based on the outer shape data of the vehicle X created by the outer shape creation unit 442 according to the discrimination model 451 (vehicle type discrimination model) stored in the storage unit 45. That is, the discrimination model 451 (discrimination model M), which is an algorithm recorded as data in the storage unit 45, is loaded and executed on the outer shape data of the vehicle X to discriminate the vehicle type of the vehicle X.

[0042] The discrimination model M (vehicle type discrimination model) is a classifier that, when input with the external shape data u of vehicle X, discriminates which of the pre - determined n types of vehicle types C1 to Cn as classes the external shape data u belongs to, that is, which class Ck (1 ≤ k ≤ n, k is a natural number) the external shape data belongs to. The discrimination model M can also be said to be a function that outputs one class Ck as the output w when the external shape data u is input. That is, the discrimination model M is a discrimination model (discrimination algorithm) with the external shape data u of vehicle X as the explanatory variable and the vehicle type information as the objective variable w.

[0043] Also, as schematically shown in FIG. 5, in the discrimination model M, when the external shape data u of vehicle X is input, it may be configured to include sub - models fk that calculate the probability Pk (Pk is a real number from 0 to 1) that the external shape data u belongs to class Ck for all classes C1 to Cn. That is, it may be an algorithm applying 1vR (One vs Rest) as the discrimination model M. Note that the probability Pk may also be referred to as the prediction probability.

[0044] Here, the sub - model fk may be a sub - model generated by a known machine learning or statistical method that outputs the prediction probability Pk, which is an index of the probability that the vehicle type belongs to class Ck, when the external shape data u is input. For example, logistic regression, support vector machine (SVM), or a neural network may be applied to the sub - model fk.

[0045] In FIG. 5, it is exemplified that the probability that the external shape data u is determined to belong to class C1 is 0.01, for C2 it is 0.01, and for C3 it is 0.96. Thus, in the discrimination model M, when the external shape data u of vehicle X is input, a list (vector) of such prediction probabilities Pk, v = {P1, P2, ···, Pn}, is calculated internally by the set of sub - models fk.

[0046] Then, inside the discrimination model M, the sub-function q extracts and outputs, from the list of prediction probabilities Pk, the index of the prediction probability Pk with the maximum prediction probability Pk, that is, the class Ck, under the condition that the prediction probability Pk exceeds the threshold value Pth. The output of the sub-function q corresponds to the final output of the discrimination model M. If there is no class Ck for which the prediction probability Pk exceeds the threshold value Pth, the discrimination model M outputs an error.

[0047] The structure of the discrimination model M shown in FIG. 5 is an example, and other known models can be used as long as they perform the same output for the external shape data u. For example, the algorithm for calculating the list (vector) v of prediction probabilities Pk from the external shape data u may not be divided into sub-models.

[0048] The storage unit 45 of the car wash machine 2 stores a change procedure database 452, which is a database regarding change information for changing a predetermined procedure of the car wash machine 2 at a specific part of a specific vehicle type for a specific vehicle type. The specific vehicle type is a vehicle type not suitable for general course washing. For example, in the brush control in general course washing selected in the car wash machine 2, it is a vehicle type equipped with parts that may be damaged due to the shape of the parts or the weakness of the mounting strength of the parts. Examples of such parts include side mirrors, antennas, rear bumpers, etc.

[0049] Alternatively, it may be a vehicle type equipped with parts that are difficult to be reflected in the external shape data of the vehicle detected by the vehicle shape detection sensor 52. Examples of such parts include fenders, door opening and closing handles, front grilles, rear garnishes, etc. The side under mirror, which is an auxiliary side mirror, is also a part that is difficult to be reflected in the external shape data, which is the silhouette data from the side of the vehicle. Alternatively, it may be a vehicle type equipped with parts attached below the detection lower limit height of the vehicle shape detection sensor 52. Examples of such parts include side spoilers, lower parts of front and rear bumpers, door side moldings, splash guards (mud guards), etc.

[0050] Alternatively, the specific vehicle type may be a vehicle type that includes components with a shape different from that of a normal vehicle type at a low position where it is difficult to grasp the shape with the vehicle shape detection sensor 52. Examples of such components include fenders and rear bumpers. Alternatively, it may be a vehicle type equipped with components having a characteristic shape, and it is preferable that brushing, rinsing, or drying at the position of the component is performed by special control. Further, the specific part means a part where a component that may be damaged as described above or a component having a characteristic shape is mounted in a specific vehicle type.

[0051] The change procedure database 452 stores change information for the specific part of the specific vehicle type. The change information includes a specific distance associated with each specific vehicle type and information indicating a change procedure for controlling each part of the car wash machine 2 at the specific distance. The specific distance is the distance from the start to the end of the change of a predetermined car wash procedure. That is, the vehicle X is washed by the changed procedure at the specific distance.

[0052] The specific distance is defined, for example, with reference to the front end position of the vehicle (as zero). Since the change information includes the specific distance and information indicating the change procedure for controlling each part of the car wash machine 2 at the specific distance, the control unit 44 can refer to the change information and change a predetermined procedure at the specific distance.

[0053] Here, the change procedure includes, for example, a procedure related to the avoidance control of the brush (top brush 14, side brush 16, or rocker brush 18) at the specific part, a procedure related to the change of the cleaning intensity by the brush, or a procedure related to the change of the rinsing amount. By instructing a change in the control of each part included in the car wash machine 2 in the change procedure, the control unit 44 can control each part to change a predetermined procedure at the specific distance.

[0054] The change procedure database 452 may store change information in a table as shown in Table 1 below, for example. In Table 1 below, the change procedure start position and the change procedure end position indicate the distance (mm) from the tip position of the vehicle X. In Table 1 below, the distance from the change procedure start position to the change procedure end position corresponds to the specific distance. TB in Table 1 means top brush.

[0055]

Table 1

[0056] Note that in FIGS. 1 and 2, for simplicity of illustration, the illustration of each device for washing the vehicle X provided in the car wash machine body 4 may be omitted. In addition, each device provided in the car wash machine body 4 shown in FIG. 1 is merely an example, and the car wash machine body 4 may be provided with devices for washing the vehicle X including conventionally known configurations in addition to the above-described devices, and devices for assisting the washing on the frame 8 or the ceiling portion 10.

[0057] <Flow of the car wash operation> FIG. 4 is a flowchart showing an example of the control of the control unit 44. In the flowchart shown in FIG. 4, the car wash control unit 441 controls the drive unit 4X to start the relative movement of the car wash machine body 4 with respect to the vehicle X, and the description starts from the time when the cleaning unit 4Y is controlled to start the car wash according to a predetermined procedure.

[0058] When the car wash is started, the outer shape creation unit 442 acquires information necessary for creating outer shape data from each part of the sensor system 7 and starts creating the outer shape data (step S1). Specifically, the outer shape creation unit 442 creates outer shape data using the detection signals of the respective photoelectric switches output by the vehicle shape detection sensor 52 and the travel count data output by the distance sensor 71. Note that the outer shape creation unit 442 attaches the creation time to the created outer shape data at an appropriate timing, that is, performs a timestamp, and stores it in the storage unit 45 as the outer shape data log 453.

[0059] When the vehicle type discrimination unit 443 passes a predetermined vehicle type discrimination point (S2: YES), it discriminates the vehicle type of the vehicle X by executing a discrimination model 451, which is an algorithm recorded as data in the storage unit 45, based on the external shape data created by the external shape creation unit 442. The vehicle type discrimination unit 443 further discriminates whether the vehicle type discriminated as the vehicle type of the vehicle X corresponds to a specific vehicle type (S3). Note that when an error is output as a result of the execution of the discrimination model 451 by the vehicle type discrimination unit 443, it is determined that the vehicle X does not correspond to the specific vehicle type.

[0060] The predetermined vehicle type discrimination point is a point at which the control unit 44 executes vehicle type discrimination, which is set in advance, and may be defined by, for example, a specified value of the travel count data output by the distance sensor 71. Alternatively, it may be defined as the time point when the detection signals of the respective photoelectric switches output by the vehicle shape detection sensor 52 satisfy a specific condition. The time point when the specific condition is satisfied may be, for example, the time point when the amount of change in the vehicle height data decreases or the vehicle height data becomes constant in the change over time of the vehicle height data indicated by the detection signal.

[0061] The time point when the amount of change in the vehicle height data decreases or the vehicle height data becomes constant may indicate the time point when the position measured by the vehicle shape detection sensor 52 has passed the front glass of the vehicle X and reached the roof of the vehicle X. That is, the control unit 44 discriminates whether the vehicle X is a specific vehicle type during the car wash of the vehicle X. Since the control unit 44 can discriminate whether the vehicle X is a specific vehicle type during the car wash of the vehicle X, it may not be necessary to provide a sensor for discriminating the vehicle X outside the car wash machine main body 4. In addition, since it is possible to discriminate whether the vehicle X is a specific vehicle type at a predetermined vehicle type discrimination point, the time required for vehicle type discrimination can be shortened.

[0062] When the vehicle type determination unit 443 determines that the vehicle X corresponds to a specific vehicle type (S3: YES), the car wash control unit 441 refers to the change procedure database 452 stored in the storage unit 45, and controls each part of the cleaning unit 4Y to change a predetermined car wash procedure based on the change information for the corresponding specific vehicle type (step S4). That is, when the vehicle X is a specific vehicle type, the control unit 44 performs car washing according to a predetermined car wash procedure changed according to the change information. When the vehicle X is not a specific vehicle type, car washing is performed according to the above-described predetermined procedure, that is, the basic procedure.

[0063] When the vehicle type determination unit 443 determines that the vehicle X does not correspond to a specific vehicle type (S3: NO), the car wash control unit 441 performs car washing without changing the predetermined procedure. When the car washing according to the predetermined procedure is completed, the control unit 44 ends the control.

[0064] With the above configuration, the car wash machine 2 according to the present disclosure can determine a specific vehicle type. Further, a predetermined car wash procedure at a specific part of the determined specific vehicle type can be changed to a car wash procedure appropriate for the specific vehicle type based on the change information. Thereby, it is possible to perform cleaning that reduces the possibility of damage to parts in a specific vehicle type. In addition, for vehicle types other than the specific vehicle type, since brush avoidance is not performed more than necessary, the occurrence of a situation where the cleaning power decreases and customer satisfaction decreases is reduced.

[0065] <Monitoring device 9> The block diagram of FIG. 2 also shows the monitoring device 9 according to the present embodiment. The monitoring device 9 includes at least one camera 96, a storage unit 95 that can hold an image data log 951 in which data of images captured by the camera 96 are sequentially recorded, an input / output unit 97 for inputting and outputting various data, and a control unit 94 that controls each part of the monitoring device 9.

[0066] The camera 96 of the monitoring device 9 is installed at least near the car wash machine main body 4 and the remote panel 6 in order to monitor criminal acts against the car wash machine 2 and record the operation of the car wash machine 2. FIG. 1 shows an example of the installation position of the camera 96. The camera 96 may be installed so as to be held by the car wash machine main body 4.

[0067] For example, the camera 96 may be installed so as to be able to photograph the vehicle X entering the car wash machine main body 4, the vehicle X being washed, the car wash machine main body 4, and the vehicle X exiting the car wash machine main body 4. The number of cameras 96 provided in the monitoring device 9 is not limited to one. The monitoring device 9 may be configured to appropriately record images during the period from when the car wash machine 2 accepts the washing of the vehicle X to when the vehicle X exits for the above purposes. In this case, the monitoring device 9 may be configured to receive a signal regarding the operation of the car wash machine 2 from the car wash machine 2 and control the operation of the monitoring device 9 based on the information of the signal.

[0068] Also, the monitoring device 9 may be configured to record an image when a person approaching the car wash machine main body 4 or the remote panel 6 is detected. In this case, the monitoring device 9 may be provided with a proximity sensor that senses the approach of an object. Alternatively, the monitoring device 9 may be configured to record images constantly.

[0069] The control unit 94 of the monitoring device 9 controls the camera 96 in this way, and attaches the shooting time to the data of the image captured by the camera 96, and accumulates it in the storage unit 95 of the monitoring device 9 as an image data log 951. All or the required part of the image data log 951 stored in the storage unit 95 can be transferred to an external device through the input / output unit 97 as appropriate.

[0070] The input / output unit 97 includes a communication interface for communicating with an external device through a network or directly. A commercially available surveillance camera system may be applied to such a monitoring device 9. Alternatively, the car wash machine 2 may be configured to incorporate the functions of such a monitoring device as a part of its functions.

[0071] <Configuration of the discrimination model generation device 100> The discrimination model generation device 100 according to one aspect of the present disclosure will be described below. The discrimination model generation device 100 is a device that generates a discrimination model M (vehicle type discrimination model) using the external shape data and image data accumulated by the above-described car washer 2 and monitoring device 9.

[0072] FIG. 6 is a block diagram showing the configuration of the discrimination model generation device 100 according to the present embodiment. The discrimination model generation device 100 includes a control unit 140, a storage unit 150 (storage device), and an input / output unit 170. The control unit 140 executes required calculations and controls each part of the discrimination model generation device 100. In the discrimination model generation device 100, the control unit 140 includes at least functional blocks of a data acquisition unit 141, a discrimination model execution unit 142, a vehicle type information acquisition unit 143, a teacher data generation unit 144, a discrimination model generation unit 145, and a discrimination model output unit 146.

[0073] The storage unit 150 is a memory that stores various types of information. In the discrimination model generation device 100, the storage unit 150 may store information such as an image data log VL, an external shape data log PL, teacher data TD, and the discrimination model M. The input / output unit 170 is a functional block for exchanging information with an external device. The input / output unit 97 includes a communication interface for communicating with an external device through a network or directly.

[0074] The discrimination model generation device 100 is typically realized by a general-purpose computer. This computer includes, for example, one or more processors and a computer-readable recording medium storing the above program. Then, in the above computer, the above object is achieved by the above processor reading and executing the above program from the above recording medium.

[0075] As the above-mentioned processor, for example, a CPU (Central Processing Unit) can be used. As the above-mentioned recording medium, in addition to "non-transitory tangible media" such as ROM (Read Only Memory), magnetic disks, cards, semiconductor memories, programmable logic circuits, etc. can be used. Further, it may further include a RAM (Random Access Memory) for expanding the above-mentioned program.

[0076] <Generation of Discrimination Model> The discrimination model M generated by the discrimination model generation device 100 is the above-mentioned discrimination model used by the car wash machine 2. The discrimination model generation device 100 generates the discrimination model M using a known machine learning method or statistical method with the set of vehicle type information and its external shape data for a vehicle X with a known vehicle type as the teacher data.

[0077] FIG. 7 shows a flowchart representing the procedure for generating the discrimination model M using the discrimination model generation device 100. Hereinafter, with reference to FIG. 7, the procedure for generating the discrimination model M will be described.

[0078] Step S11: The data acquisition unit 141 of the discrimination model generation device 100 acquires at least a part of the external shape data log 453 stored in the storage unit 45 of the car wash machine 2 and stores it in the storage unit 150 as the external shape data log PL. Further, the data acquisition unit 141 acquires at least a part of the image data log 951 stored in the storage unit 95 of the monitoring device 9 and stores it in the storage unit 150 as the image data log VL. The car wash machine 2 and the monitoring device 9 for which the data acquisition unit 141 acquires these data are not limited to a single one, and may be a plurality.

[0079] When the car wash machine 2 or the monitoring device 9 is connected to the communication network, the data acquisition unit 141 may access the car wash machine 2 or the monitoring device 9 from the input / output unit 170 via the communication network to acquire these data.

[0080] When the car washer 2 or the monitoring device 9 is not connected to the communication network, a support staff member who performs maintenance on the car washer 2 may go to the car washer 2 or the monitoring device 9 and temporarily acquire these data from the car washer 2 or the monitoring device 9 onto an electronic device such as a portable storage device or a portable computer. In this case, the data acquisition unit 141 acquires these data from the car washer 2 or the monitoring device 9 via the electronic device.

[0081] Step S12: Next, a candidate for the external shape data to be added to the teacher data TD is selected. When a large number of data have already been accumulated as the teacher data TD and a discrimination model M using the teacher data has been generated and stored in the storage unit 150, the teacher data that should be particularly added is data regarding a vehicle type (class) that was not included in the already accumulated teacher data TD. Extraction of data regarding a new vehicle type for such a discrimination model M can be performed as follows.

[0082] For the external shape data in the external shape data log PL, the discrimination model execution unit 142 applies the already constructed discrimination model M to perform discrimination. As a result, when the discrimination model M outputs an error for a certain external shape data, since there is no class Ck of the vehicle type that matches in the already constructed discrimination model M, the teacher data generation unit 144 extracts the said external shape data as data regarding a new vehicle type.

[0083] Note that when the car washer 2 stores the generated external shape data as the external shape data log 453 and attaches and stores the discrimination result by the discrimination model 451 executed by the vehicle type discrimination unit 443 of the car washer 2 to the external shape data, the execution of the discrimination model M by the discrimination model execution unit 142 can be omitted. In this case, the teacher data generation unit 144 extracts, as data regarding a new vehicle type, the external shape data with the discrimination result of error from among the external shape data included in the external shape data log PL.

[0084] Step S13: The vehicle type information acquisition unit 143 extracts, from the image data stored in the image data log VL, the captured image by the monitoring device 9 when the car wash machine 2 calculates the external shape data, by referring to the time stamp of the external shape data and the time stamp of the image data.

[0085] The vehicle type information acquisition unit 143 displays an image based on the image data on the display, and requests the operator to input the vehicle type of the vehicle in the image. When the operator performs an operation such as using the keyboard of the input / output unit 170 to input the vehicle type of the vehicle in the image into the discrimination model generation device 100, the vehicle type information acquisition unit 143 acquires the vehicle type information regarding the external shape data.

[0086] Step S14: The teacher data generation unit 144 adds the pair of the extracted external shape data and the vehicle type information acquired by the vehicle type information acquisition unit 143 to the teacher data TD stored in the storage unit 150. That is, the teacher data generation unit 144 associates the image of the vehicle X captured by the camera 96 of the monitoring device 9 with the vehicle shape data based on the output data of the vehicle shape detection sensor 52 that detected the vehicle X, and stores it in the storage unit 150 of the discrimination model generation device 100 as the teacher data TD. Steps S12 to S14 are repeatedly executed as appropriate.

[0087] Step S15: The discrimination model generation unit 145 generates a discrimination model with the external shape data u of the vehicle X as the explanatory variable and the class Ck of the vehicle type information as the target variable w, using the teacher data TD stored in the storage unit 150, and stores it in the storage unit 150. In that case, the discrimination model generation unit 145 defines the classes within the range of vehicle types including the new vehicle type included in the teacher data TD.

[0088] Step S16: The discrimination model output unit 146 outputs the newly generated discrimination model M to the car wash machine 2 through the input / output unit 170. The data transfer between the discrimination model generation device 100 and the car wash machine 2 can be performed in the same manner as described in step S11. The car wash machine 2 stores the discrimination model M received from the discrimination model generation device 100 in the storage unit 45 of the car wash machine 2 as the discrimination model 451.

[0089] Also, the above step S12 may be changed to the following method. The following method can also be applied at an initial stage where the discrimination model M has not yet been constructed. When steps S12 to S14 are appropriately repeated and executed, in step S12, the teacher data generation unit 144 sequentially selects the external shape data in the external shape data log PL.

[0090] As described above, by using the discrimination model generation device 100, it becomes possible to obtain the external shape data associated with the vehicle type from a large amount of external shape data of the vehicle X created by the car wash machine 2, and the discrimination model M used in the car wash machine 2 can be efficiently generated. Also, even when a new vehicle type appears on the market, it is possible to easily reconstruct the discrimination model M that can discriminate including the new vehicle type.

[0091] <Modification Example 1> The vehicle shape detection sensor 52 of the car wash machine 2 does not necessarily have to be held by the car wash machine main body 4. It may be installed in front of the car wash machine main body 4 (in the positive x-axis direction) so as to be fixed to the ground and configured to detect the vehicle X entering the car wash machine main body 4. In this case, in order to specify the position in the front-rear direction of the vehicle X in the external shape data, instead of the distance sensor 71 using the encoder 13 of the car wash machine main body 4, a distance measurement sensor for detecting the position of the vehicle X entering the car wash machine 2 in the x-axis direction may be fixedly installed on the ground.

[0092] When the vehicle shape detection sensor 52 is installed in front of the car wash machine main body 4, the outer shape data can be acquired at an earlier stage compared to the case where it is held by the car wash machine main body 4.

[0093] <Modification Example 2> When the car wash machine 2 is constantly connected to the network, at least any one of the discrimination model 451, the change procedure database 452, and the outer shape data log 453 may be stored on a server constantly connected to the network. Further, when the monitoring device 9 is constantly connected to the network, the image data log 951 may be stored on a server constantly connected to the network.

[0094] In such a case, the data of the discrimination model 451 and the change procedure database 452 may be obtained by the car wash machine 2 accessing the server when the car wash machine 2 uses them. In Modification Example 2, by using the server, the discrimination model generation device 100 can update the discrimination model M earlier.

[0095] 〔Summary〕 (1) The model generation method according to Aspect 1 of the present disclosure is vehicle type information indicating the vehicle type of a vehicle to be washed by a car wash machine main body that washes the vehicle while making a relative movement with respect to the vehicle in the front-rear direction of the vehicle, and includes a step of acquiring vehicle type information obtained from an image of the vehicle taken by a camera that captures at least a part of the vehicle, and a sensor for detecting the vehicle to be washed, the sensor having a plurality of optical axis sensors arranged parallel to each other along the optical axis and detecting the presence or absence of an object on the optical axis, and acquiring vehicle shape data of the vehicle generated based on output data of the sensor that makes a relative movement with respect to the vehicle in the front-rear direction of the vehicle, and generating a vehicle type discrimination model using the acquired vehicle shape data as an explanatory variable and the acquired vehicle type information as an objective variable.

[0096] (2) The model generation method according to Embodiment 2 of the present disclosure further includes a step of associating the image of the vehicle captured by the camera with the vehicle shape data based on the output data of the sensor that detected the vehicle, and storing the associated data in a storage device, in the model generation method according to Embodiment 1 above.

[0097] (3) The car wash according to Embodiment 3 of the present disclosure generates the vehicle shape data for the vehicle to be washed, discriminates the vehicle type of the vehicle using the vehicle type discrimination model generated by the model generation method of Embodiment 1 or 2 above, and executes the washing of the vehicle according to the result of the vehicle type discrimination.

[0098] The present disclosure is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present disclosure.

Explanation of Reference Numerals

[0099] X Vehicle 2 Car Wash 4 Car Wash Main Body 4X Driving Unit 4Y Cleaning Unit 14 Top Brush 16 Side Brush 18 Locker Brush 6 Remote Panel 7 Sensor System 52 Vehicle Shape Detection Sensor (Sensor) 71 Distance Sensor 44 Control Unit 45 Storage Unit 47 Input / Output Unit 9 Monitoring Device 94 Control Unit 95 Storage Unit 96 Camera 97 Input / Output Unit 100 Discrimination Model Generation Device 140 Control Unit 150 Storage Unit (Storage Device) VL Image Data Log PL Appearance Data Log TD Teacher Data M Discrimination Model (Vehicle Type Discrimination Model) 170 Input / Output Unit

Claims

A car wash machine body that washes the vehicle while relatively moving with respect to the vehicle in the longitudinal direction of the vehicle, A camera that photographs the vehicle, A sensor for detecting the vehicle to be washed, having a plurality of optical axis sensors arranged parallel to each other along the optical axis and detecting the presence or absence of an object on the optical axis, wherein the optical axes of some of the optical axis sensors are arranged so as to cross the vehicle, and the sensor relatively moves with respect to the vehicle in the longitudinal direction of the vehicle, A storage unit that stores a vehicle type discrimination model constructed using vehicle type information indicating the vehicle type of the vehicle, obtained based on at least a part of the image of the vehicle photographed by the camera, and vehicle shape data of the vehicle generated based on the output data of the sensor, with the vehicle shape data as an explanatory variable and the vehicle type information as an objective variable, A control unit, and When the control unit executes the washing of the vehicle, Based on the output data of the sensor, generates the vehicle shape data for the vehicle to be washed, Uses the vehicle type discrimination model to discriminate the vehicle type of the vehicle, A car wash machine that executes the washing of the vehicle based on the output data of the sensor for the vehicle and according to the result of the discrimination of the vehicle type.

Images