JP2020067823A - Vehicle space full/vacancy detection system of parking lot - Google Patents

Abstract

To provide a vehicle space full/vacancy detection system of a parking lot capable of highly precisely determining the existence of a vehicle from an image photographed by an image photographing unit, and, if necessary, expanding an application range of vehicle space full/vacancy detection by using licence number recognition based on the same image photographed by the image photographing unit.SOLUTION: A vehicle space full/vacancy detection system includes: an image photographing unit that photographs an image of a vehicle in a vehicle space where the vehicle is parked; a license number recognition unit that recognizes the license number of the vehicle using the photographed image; and a vehicle space full/vacancy detection unit that detects the vehicle space full/vacancy state from the photographed image.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a device for detecting fullness (whether a vehicle is parked or empty) in a vehicle compartment (vehicle parking section) of a parking lot, and more particularly to a device for performing detection by image recognition.

  At present, a system using a loop coil sensor and a lock plate (flap) is generally used for the vacancy management of a parking lot, especially a coin parking lot. This is a mechanism to check if the passenger compartment is full with a loop coil sensor and to prevent the vehicle before adjustment from exiting with a lock plate.

  However, in the lock plate method, an accident may occur in which the user stumbles on the lock plate and injuries the user, the lock plate fails in snowy areas, the lock plate does not lift due to the weight of the snow, and the snow blower cannot be used. There are also problems such as being forced to remove snow manually.

  Therefore, in recent years, instead of a lock plate, a pole equipped with a camera has been set up at the end of each passenger compartment, and it is not locked so that the vehicle cannot exit, but by recognizing the number from the camera and linking it with the adjustment data, A mechanism (lockless parking lot system) that can identify a vehicle that has left without leaving afterwards has been put into practical use.

  For example, in Patent Document 1, a vehicle number (vehicle number) is recognized from an image captured by a camera provided for each parking compartment, stored, and a fee is paid at a parking lot re-entered when the vehicle is unpaid. The claimed technical idea is disclosed.

  According to the lockless (or flapless) parking lot system of this technical idea, the fullness detection of the vehicle itself is performed by the loop coil sensor, and the exit from the vehicle without checking out is prevented by the vehicle number recognition. .

  However, the loop coil sensor as disclosed in Patent Document 1 is used widely because it is used to determine the presence of a vehicle using impedance, and it is possible to perform the determination accurately at low cost. To do this, civil engineering work is required, and this cost is a heavy burden on businesses.

  On the other hand, as a parking lot operator, it is important to install or remove the parking lot equipment in a short period of time, so when installing a pole equipped with a vehicle number recognition camera, at the same time when the vehicle is full If it can be used for detection (determination of the presence of a vehicle), equipment such as a loop coil becomes unnecessary, which is a very effective solution.

  For example, Patent Document 2 discloses a technical idea in which an image recognition camera is installed in each passenger compartment and the same vehicle is continuously parked if the main part of the license plate read at regular intervals does not change. Has been done.

  However, in Patent Document 2, when the vehicle has entered and exited the passenger compartment, the vehicle number is hidden and it is extremely difficult to determine whether or not the vehicle is continuously stopped.

JP, 2007-317098, A JP, 2008-41398, A

In this way, a method of using a loop coil to detect whether or not a vehicle exists in the passenger compartment,
There is a method of recognizing the car number, but each has its problems.

  In view of the above problems, the present invention determines whether or not a vehicle exists with high accuracy from an image captured by an image capturing unit, and further, if necessary, uses an image captured by the same image capturing unit. An object of the present invention is to provide a vehicle interior fullness detection device for a parking lot, which can widen the application range of vehicle interior fullness detection using vehicle number recognition.

In order to solve such a problem, the vehicle interior fullness detection device of the first aspect of the present invention,
An image capturing unit that captures an image of the vehicle in a passenger compartment where the vehicle is parked,
A vehicle interior fullness detection unit for detecting a fullness state of the vehicle interior from the photographed image.

  In this way, from the image captured by the image capturing unit such as a camera, it is possible to determine whether the vehicle interior, that is, a compartment such as a parking lot where the vehicle can be parked is full, that is, whether the vehicle exists in the vehicle interior. Since the detection is performed, a lockless parking lot can be realized without requiring any auxiliary equipment such as a loop coil sensor for detecting the fullness of the passenger compartment.

  By doing so, the vehicle number (vehicle number) on the license plate of the vehicle parked in the compartment where the vehicle can be parked, such as the parking lot, from the image captured by the image capturing unit such as the same camera. ) And the fullness of the passenger compartment, that is, whether or not the vehicle is present, is detected, and there is no need for ancillary equipment such as a loop coil sensor for detecting the fullness of the passenger compartment. A parking lot can be realized.

  Further, the vehicle interior fullness detection device of the second aspect of the present invention is characterized by further including a vehicle number recognition unit that recognizes the vehicle number of the vehicle from the photographed image.

  In this way, it is possible to recognize the fullness of the vehicle compartment and the recognition of the vehicle number (vehicle number) on the license plate of the vehicle parked in the vehicle compartment from the images captured by the image capturing unit such as the same camera. Since both of them are performed, it is possible to improve the accuracy of the detection of the fullness of the passenger compartment, and it is expected to realize an applied function that combines vehicle detection and vehicle number recognition.

  Further, a vehicle interior fullness detection device according to a third aspect of the present invention is the vehicle interior fullness detection device according to the first or second aspect, in which the past data and the present data of the vehicle interior fullness detection unit are compared. It is characterized in that the identity of the vehicle parked in the vehicle compartment is determined by comparing.

  Here, the past data and the present day may be the ones that have the same characteristics even if the details are not the same.

  By doing so, for example, even if the trunk or the rear hatch of the vehicle is opened, it is possible to determine that they are the same vehicle, which is extremely effective in operating the parking lot.

  Further, the vehicle interior fullness detection device of a fourth aspect of the present invention is the vehicle interior fullness detection device of the first or second aspect, wherein the vehicle interior fullness detection unit detects movement of the vehicle. Is characterized in that it is not determined that there is a vehicle in the passenger compartment.

  That is, the moving vehicle is prevented from being erroneously determined to be parked, which improves the accuracy of parking lot management.

  Further, a vehicle interior fullness detection device of a fifth aspect of the present invention is the vehicle interior fullness detection device of the first or second aspect, which transmits the information when the image capturing section is difficult to capture. It is characterized by doing.

  By doing this, for example, when the image capturing unit is in a non-functioning state, such as when the image capturing unit is shielded by mischief or maliciously, or when the image capturing unit or the power supply for the image capturing unit is broken, By notifying the information to the parking lot control unit or the like that controls the entire parking lot, it becomes possible to operate the parking lot properly.

  Further, a vehicle interior fullness detection device according to a sixth aspect of the present invention is the vehicle interior fullness detection device according to the first or second aspect, wherein the vehicle interior fullness detection unit uses a result of another sensor. Then, the parameters are automatically learned to improve the detection accuracy.

  That is, other than the vehicle interior fullness detection unit based on the image, for example, by using the result of the loop coil sensor or the vehicle number recognition, automatic learning by a method such as deep learning is performed at the installed location, and the parameters are set. To improve the detection accuracy.

  Alternatively, an image collecting function may be connected to the image capturing unit, and the image information and the detection result of the passenger compartment fullness detection unit may be stored together and learned at a later date or in another place.

  In either case, it can be expected that images and sensor detection results will be accumulated and that the passenger compartment fullness detection performance will be dramatically improved.

  In addition, a vehicle interior fullness detection device of a seventh aspect of the present invention, in the vehicle interior fullness detection device of the first aspect, by combining with other sensors, the detection performance of the vehicle interior fullness detection unit. It is characterized by improving.

  That is, even when the accuracy of the detection of the vehicle cabin fullness by the image is low, the loop coil sensor (difficult to detect in a vehicle with a high vehicle height), the infrared sensor (difficult to distinguish from a person), the ultrasonic sensor (distance to a person) It is difficult to distinguish between each other), and an image sensor with a bird's-eye view from above (difficulty with respect to vehicle overlap and blockage), or two or more sensors are used together with the vehicle interior full sky detection. By doing so, it is possible to improve the detection performance (correct answer rate) regarding the presence or absence of the vehicle in the vehicle interior.

  Also, the vehicle interior fullness detection device of the eighth aspect of the present invention is the vehicle interior fullness detection device of the first or second aspect, wherein the vehicle interior fullness detection unit recognizes a feature of the vehicle It is characterized by detecting the presence or absence of.

  That is, it is possible to determine that a target, such as a large vehicle or a vehicle with a high vehicle height, which cannot be detected by the loop coil sensor, is a vehicle by grasping the entire outer shape of the vehicle. , It is possible to detect when the passenger compartment is full.

  Further, a vehicle interior fullness detection device of a ninth aspect of the present invention is the vehicle interior fullness detection device of the first or second aspect, wherein the vehicle interior fullness detection unit is additional information other than the presence or absence of the vehicle. Is detected.

  That is, in addition to the presence / absence of the vehicle, the movement of the vehicle can be used to detect the entry / exit of the vehicle into / from the passenger compartment, determine the vehicle type and the paint color of the vehicle, and determine the presence / absence of a person (driver, etc.) in the vehicle. Information may be detected. In this way, the vehicle interior fullness detection device can be applied to a wider range.

  Further, a vehicle interior fullness detection device according to a tenth aspect of the present invention is the vehicle interior fullness detection device according to the first or second aspect, wherein the image capturing section or another image capturing device is a vehicle. It is characterized in that the presence or absence and / or the degree of damage of the vehicle is determined from an image taken while the vehicle is parked in the vehicle or when entering or leaving the vehicle.

  This makes it easy to identify the time of damage and the degree of damage, especially in the case of car sharing, and it is possible to take appropriate measures.

  Further, the vehicle interior fullness detection device of an eleventh aspect of the present invention is the vehicle interior fullness detection device of the second aspect, wherein the vehicle interior fullness detection is performed regardless of whether or not the vehicle number recognition unit has succeeded in recognition. It is characterized by using the detection result of the copy.

  By doing this, even if the license plate or the part of the license number on the license plate cannot be recognized because it is dirty, damaged, or concealed, the passenger compartment is full. Information can be obtained on the status (whether or not there is a vehicle), and the parking lot can be operated in accordance with the information, so the utilization can be expanded.

  At that time, the reliability of the detection of the full state of the passenger compartment may be increased by also using the vehicle number recognition result.

  According to the vehicle interior fullness detection device for a parking lot of the present invention, whether or not a vehicle is present in the vehicle interior is determined with high accuracy from an image captured by an image capturing unit such as a camera. It is possible to detect the fullness of the passenger compartment without using the vehicle detection means described above, and it is possible to install and operate the parking lot at low cost. Further, in combination with the vehicle number recognition for recognizing the vehicle number from the images captured by the same image capturing unit, various synergistic effects can be obtained.

FIG. 1 is a block diagram of a passenger compartment fullness detection device according to an embodiment of the present invention. It is a perspective view of a vehicle interior fullness detection device according to an embodiment of the present invention. 4 is a flowchart of the operation of the device according to the embodiment of the present invention. 4 is a flowchart of the operation of the device according to the embodiment of the present invention. 3 is a flowchart of a vehicle compartment fullness detection unit of the device according to the embodiment of the present invention. It is explanatory drawing of the image of the passenger compartment fullness detection apparatus which concerns on one Embodiment of this invention. It is explanatory drawing of the image of the passenger compartment fullness detection apparatus which concerns on one Embodiment of this invention. 5 is another flowchart of the operation of the device according to the embodiment of the present invention. 5 is another flowchart of the operation of the device according to the embodiment of the present invention.

  Hereinafter, a passenger compartment fullness detection device according to an embodiment of the present invention will be described with reference to the drawings. In the following, the range necessary for the explanation to achieve the object of the present invention is schematically shown, and the range necessary for the explanation of the relevant part of the present invention will be mainly explained, and the explanation is omitted. A known technique is used.

  FIG. 1 is a block diagram of a vehicle interior fullness detection device 1 according to one embodiment of the present invention, and FIG. 2 is a perspective view of the vehicle interior fullness detection device 1 according to one embodiment of the present invention.

  The vehicle interior fullness detection device 1 has at least an image capturing unit 11, a vehicle interior fullness detection unit 12, and a vehicle number recognition unit 13, and further includes a control unit 14 including a storage unit and a display unit as necessary. You may have 15, the audio | voice guidance part 16, the operation part 17, etc.

  Here, as shown in FIG. 2, the vehicle compartment fullness detection device 1 is located at the rear of the vehicle entry direction of each vehicle compartment 21 in a parking lot provided with a plurality of vehicle compartments 21 divided by a partition line 20. The main elements are built in the provided pole 22.

  A car stop 23 is provided at a rear portion of the vehicle interior 21 to prevent the vehicle from entering a predetermined position to a depth and damaging surrounding people or the pole 22.

  The pole 22 has a substantially rectangular parallelepiped column shape and extends in the up-down direction, and an image capturing unit (camera) 11 is built in the upper portion thereof.

  The image capturing unit 11 may be any well-known unit as long as it can capture a still image or a moving image, and its capturing range may be at least capable of capturing the front passenger compartment 21 in which the pole 23 is installed.

  The image capturing units 11 are preferably provided in the respective vehicle compartments 21, but it is also possible to capture a plurality of vehicle compartments 21 with one image capturing unit 11 in terms of the layout of the parking lot, costs, and the like. It may be effective in.

  Although not shown in FIG. 2, the vehicle cabin fullness detection unit 12 and the vehicle number recognition unit 13 may be built in the pole 22 or may be provided outside the pole 22. In that case, the vehicle compartment fullness detection unit 12 and the vehicle number recognition unit 13 for a plurality of vehicle compartments may be installed together with the control unit 14 or may be functionally integrated. .

  The vehicle interior fullness detection unit 12 detects the presence or absence of a vehicle in the vehicle interior from the image captured by the image capturing unit 11, and is preferably in the form of a program incorporated in the vehicle interior fullness detection device 1. .

  The vehicle number recognition unit 13 recognizes a vehicle number (number information described on the license plate) attached to the vehicle from the image captured by the image capturing unit 11, and is preferably the passenger compartment fullness detection device 1 It is a form of the program built in.

  The control unit 14 controls the entire passenger compartment fullness detection device 1, and preferably has a CPU, a memory, a clock function, and the like.

  The display unit 15 is provided on the front side of the pole 22 and below the image capturing unit 11, and can display various information. The configuration may be a well-known display such as an LED or a plasma display, or may be a fixed information display and a correspondence such as a lamp.

  The voice guidance unit 16 is provided on the front side of the pole 22 and below the display unit 15, and can guide various information. The configuration may be a well-known speaker or the like.

Further, although not shown, the operation unit 17 including a keyboard or the like may be provided on the pole 22.
In some cases, it is necessary for instructions to display guidance or to settle fees. The operation unit 17 may be in the form of a touch panel provided on the surface of the display unit 15.

  In addition, the passenger compartment fullness detection device 1 may include the communication interface unit 18 and may be connected to the external charge settlement system 100 or the monitoring system 200 via a line or the like.

  Next, the operation of the vehicle interior fullness detection device according to the embodiment of the present invention will be described. Here, the case where the vehicle interior fullness detection device does not have a vehicle number recognition unit will be described.

  FIGS. 3 and 4 are flowcharts of the operation of the vehicle interior fullness detection device according to the embodiment of the present invention. FIG. 3 shows a state in which the vehicle interior is empty, and FIG. 4 starts from a state in which the vehicle interior is full. It is a flow to do.

  It is assumed that the specific passenger compartment 21 in which the passenger compartment fullness detection device 1 is installed is empty. In step S01, the image capturing unit 11 continues to capture the vehicle interior 21. Here, it is assumed that the passenger compartment fullness detection unit 12 detects a vehicle in the passenger compartment from an image obtained by capturing the passenger compartment 21.

  The vehicle detection method of the vehicle detection unit 12 is as follows. FIG. 5 is a flow chart of a passenger compartment fullness detection unit according to an embodiment of the passenger compartment fullness detection apparatus of the present invention, and FIGS. 6 and 7 are flowcharts of vehicle compartment fullness detection apparatus according to one embodiment of the present invention. It is explanatory drawing of the image of.

  The vehicle detection unit 12 extracts what is considered to be an image of the vehicle from the captured image by machine learning, and in particular, deep learning using an artificial neural network.

  Here, first, classified learning data (image with a vehicle in the passenger compartment / image with no vehicle in the passenger compartment) is collected from the images captured by the image capturing unit 11 (step S11). It is desirable that the amount is large, and an image other than the image captured by the image capturing unit 11 may be used.

  Next, the learning data is divided into training data and verification data (step S12).

  With respect to this, learning and weight adjustment are performed using training data using a neural network, and the current learning situation is verified using verification data (step S13).

  By repeating this, if the error rate at the time of verification becomes smaller than the desired value, the learning ends (step S14).

  Next, a new image, for example, an image of the front part of the vehicle as shown in FIG. 6 captured by the image capturing unit 11 is input to the neural network for which learning has been completed (step S15).

  Then, this image is determined to be an image of the vehicle, and the presence of the vehicle in the vehicle compartment is detected. (Step S16).

  If the vehicle is reflected in the edge of the image, it is possible to detect that the vehicle is detected, but by learning to determine that there is no vehicle in the passenger compartment, ensure that the vehicle is in the passenger compartment. Can be detected.

  Similarly, even if the image of the rear part of the vehicle as shown in FIG. 7 is input, it is determined that this image is the image of the vehicle, and the presence of the vehicle in the vehicle compartment is detected.

  Note that the machine learning method is not limited to neural networks and deep learning, and may be machine learning of any method such as decision tree learning or support vector machine (step VM), but additional manual input is required. It may be necessary.

  Also, methods that do not rely on machine learning can be used. For example, edge extraction processing may be used. This is to consider a place where there is a sharp change in the brightness of an image as an edge, that is, a contour, and a specific method has been put to practical use.

  Returning to FIG. 3, in step S02, when the vehicle interior fullness detection unit 12 detects a vehicle in the vehicle interior, then in step 03, the control unit 14 determines the vehicle interior number, the vehicle image, the vehicle from the detection result. The information such as the characteristics and the storage time is stored in a memory and / or transmitted to an external charge settlement system 100 or the like.

  It should be noted that the information to be stored listed here is an example, and it may be more or less than this.

  If the vehicle is not detected in the vehicle compartment in step 02, the process returns to the original state and the image capturing of the vehicle compartment is continuously or intermittently continued.

  Next, a flow when the vehicle compartment is full will be described. In FIG. 4, it is assumed that the specific passenger compartment 21 in which the passenger compartment fullness detection device 1 is installed is full.

  In step S21, the image capturing unit 11 continues to capture the image of the vehicle interior 21, and the vehicle interior fullness detection unit determines whether to detect the vehicle in the vehicle interior from the captured image. (Step S22)

  Here, it is assumed that the passenger compartment fullness detection unit 12 no longer detects the vehicle in the passenger compartment from the captured image.

  In this case, it is assumed that the vehicle has moved out of the passenger compartment and has left the parking lot, and information is exchanged with the connected fee adjustment system 100 to determine whether the fee adjustment is completed. . (Step S23)

  If the payment has been made, it is regarded as a normal end and the passenger compartment is returned to an empty state, and image capturing is continued. (Step S24)

  If the charge has not been settled, the process proceeds to the abnormal process of step S26. Here, information such as the vehicle image, the characteristics of the vehicle, and the entry / exit time is stored, and the fee is separately charged, or the user collects it when the user uses it again later.

  When the detection of the vehicle in the vehicle interior is continued in step 22, it is continuously determined whether or not the vehicles are the same. (Step 25)

  Regarding the detection of the vehicle, the identity of the vehicle parked in the passenger compartment 21 is determined by comparing the past vehicle image of the passenger compartment fullness detection unit 12 and the data such as the characteristics of the vehicle with the current data.

  Here, the past data and the current data may have the same characteristics even if the details are not the same. By doing so, for example, even when the trunk or the rear hatch of the vehicle is opened, it is possible to determine that they are the same vehicle.

  In step 25, when the presence of the same vehicle can be confirmed, the process returns to the original state and the image capturing of the vehicle interior is continued.

  If it cannot be confirmed that they are the same vehicle, the abnormality processing is performed. (Step S26) In this case, it is unclear whether the same vehicle is continuously parked. Therefore, whether the detection of the vehicle has only failed, or the vehicle has been clearly replaced. Take action.

  Next, an operation when the vehicle interior fullness detection device has a vehicle number recognition unit will be described. 8 and 9 are flowcharts of the operation of the vehicle interior fullness detection device according to the embodiment of the present invention. FIG. 8 is a state in which the vehicle interior is empty, and FIG. 9 is a state in which the vehicle interior is full. It is a flow to do.

  It is assumed that the specific passenger compartment 21 in which the passenger compartment fullness detection device 1 is installed is empty. In step S31, the image capturing unit 11 continues to capture the vehicle interior 21. Here, it is assumed that the passenger compartment fullness detection unit 12 detects a vehicle in the passenger compartment from an image of the passenger compartment 21.

  The vehicle detection method of the vehicle detection unit 12 is the same as the content described above with reference to FIG.

  Returning to FIG. 8, in step S32, when the vehicle interior fullness detection unit 12 detects the vehicle in the vehicle interior, the vehicle number recognition unit 13 then recognizes the vehicle number using the image of the same vehicle. To do. (Step S33) Regarding the vehicle number recognition, various OCR technologies have been put into practical use, and these may be applied.

  If the passenger compartment fullness detection unit 12 does not detect a vehicle in the passenger compartment, the process does not proceed to step S33, and the capturing of the image of the passenger compartment is continued.

  When the vehicle number is recognized in step S33, the information is sent to the control unit 14 of the vehicle interior vacancy detection device 1, and the control unit 14 determines the vehicle number, the vehicle number (vehicle number) from the detection result. ), Vehicle images, vehicle characteristics, information such as warehousing time, are stored in a memory and / or transmitted to an external charge settlement system 100 or the like.

  It should be noted that the information to be stored listed here is an example, and it may be more or less than this.

  If the vehicle number is not recognized in step S33, the process proceeds to step S35 for abnormal processing. That is, this is because the vehicle cannot be specified because the vehicle number is not recognized although the vehicle is detected.

  In this case, even if the vehicle number is unknown, the vehicle image information and the warehousing time are stored in advance so that the payment can be made.

  Next, a flow when the vehicle compartment is full will be described. In FIG. 9, it is assumed that the specific passenger compartment 21 in which the passenger compartment fullness detection device 1 is installed is full.

  In step S51, the image capturing unit 11 continues to capture the image of the vehicle interior 21, and the vehicle interior fullness detection unit determines whether to detect the vehicle in the vehicle interior from the captured image. (Step S52)

  Here, it is assumed that the passenger compartment fullness detection unit 12 no longer detects the vehicle in the passenger compartment from the captured image.

  In this case, it is assumed that the vehicle has moved out of the passenger compartment and has left the parking lot, and information is exchanged with the connected fee adjustment system 100 to determine whether the fee adjustment is completed. . (Step S53)

  If the payment has been made, it is regarded as a normal end and the passenger compartment is returned to an empty state, and image capturing is continued. (Step S54)

  If the charge has not been settled, the process proceeds to the abnormal process of step S57. Here, information such as the vehicle number, the image of the vehicle, and the entry / exit time is stored and separately charged, or collected when the user uses the item again later.

  On the other hand, in step 52, when the detection of the vehicle in the vehicle interior is continued, the vehicle number is continuously recognized. (Step 55)

  Regarding the detection of the vehicle, it is also possible to determine the identity of the vehicle parked in the passenger compartment 21 by comparing the past data of the passenger compartment fullness detection unit 12 with the current data. Here, the past data and the current data may have the same characteristics even if the details are not the same. By doing so, for example, even when the trunk or the rear hatch of the vehicle is opened, it is possible to determine that they are the same vehicle.

  In step 55, if the vehicle number is not recognized, the vehicle exists and the vehicle number cannot be recognized, so that the abnormality processing is performed. (Step S57) In this case, it is unknown whether or not the same vehicle is continuously parked, and therefore, a countermeasure is taken according to a possible situation.

  In step 55, it is determined whether the recognized vehicle number is the same as the vehicle number recognized so far. (Step S56) If they are the same, it is considered that the parking is continued, the process returns to step S51, and the image capturing is continued.

  On the other hand, in step 55, when the recognized vehicle number is different from the vehicle numbers recognized up to that point, the abnormality processing of step 57 is performed.

  In this case, it is presumed that the vehicles have been replaced in between the image captures. Therefore, as described above, after confirming whether or not there is a charge settlement, the necessary measures will be taken.

  In the vehicle interior fullness detection device 1 of the present invention described above, when the vehicle interior fullness detection unit 12 detects the movement of the vehicle, it may be determined that the vehicle does not exist in the vehicle interior. desirable.

  That is, the moving vehicle is prevented from being erroneously determined to be parked, which improves the accuracy of parking lot management.

  Further, in the vehicle interior fullness detection device 1 of the present invention described so far, when the image capturing unit 11 has difficulty in capturing the image, the information may be transmitted.

  In this case, the image capturing unit 11 is in a non-functioning state, for example, when the image capturing unit 11 is shielded by mischief or maliciously, or when the image capturing unit 11 or a power source for the image capturing unit 11 is broken. In this case, the information can be notified to the charge settlement system 100, the monitoring system 200, a parking lot control unit (not shown) that controls the entire parking lot, or the like so that appropriate parking lot operation can be performed.

  Further, the vehicle interior fullness detection unit 12 of the present invention described so far may automatically learn parameters by using the results of other sensors (not shown) to improve the detection accuracy.

  That is, other than the vehicle interior fullness detection unit 12 using an image, for example, using a result of a loop coil sensor or a vehicle number recognition, automatic learning by a method such as deep learning is performed at a place where the parameters are set. To improve the detection accuracy.

  Alternatively, an image collecting function may be connected to the image capturing unit 11, and the image information and the detection result of the vehicle interior fullness detection unit 12 may be accumulated together and may be learned at a later date or in another place.

  Further, the vehicle interior fullness detection unit 12 of the present invention described so far may be used in combination with other sensors to improve the detection performance.

  That is, even when the accuracy of the detection of the vehicle cabin fullness by the image is low, the loop coil sensor (difficult to detect in a vehicle with a high vehicle height), the infrared sensor (difficult to distinguish from a person), the ultrasonic sensor (distance to a person) It is difficult to distinguish between each other), and an image sensor with a bird's-eye view from above (difficulty in overlapping or blocking the vehicle), or two or more of them are used together with the vehicle interior fullness detection by the image. As a result, the detection performance (correct answer rate) regarding the presence / absence of a vehicle in the passenger compartment can be improved.

  Further, the vehicle interior fullness detection unit 12 of the present invention described so far may be configured to recognize the characteristics of the vehicle and detect the presence or absence of the vehicle. That is, it is possible to determine that a target, such as a large vehicle or a vehicle with a high vehicle height, which cannot be detected by the loop coil sensor, is a vehicle by grasping the entire outer shape of the vehicle. , It is possible to detect when the passenger compartment is full.

  Further, the vehicle interior fullness detection unit 12 of the present invention described so far may detect additional information other than the presence or absence of the vehicle. That is, in addition to the presence / absence of the vehicle, the movement of the vehicle detects the entry / exit of the vehicle into / from the passenger compartment 21, the discrimination of the vehicle type, the painting color of the vehicle, and the presence / absence of a person (driver etc.) in the vehicle. The information may be detected.

  In addition, when the vehicle interior fullness detection unit 12 and the vehicle number recognition unit 13 of the present invention described so far are interlocked with each other, when it is detected that the vehicle designated in advance is correctly parked in the vehicle interior 21, the You may make it notify.

  In this way, for example, if the user who reserved the car sharing is notified of the result, it is possible to timely know that the reserved vehicle has already arrived in the designated vehicle compartment 21, and the user The convenience of is greatly improved.

  The current car share is an extension of the rental car, and it is assumed that the user will drive himself and return it at the end. However, in the future, if the technology for autonomous driving of vehicles evolves, it will be possible to automatically move to the destination and be able to drop off, it will have a meaning similar to buses and taxis, so It is expected that such needs will emerge in the future.

  In addition, it is expected that there will be a need to select various vehicle types depending on the mood and scene as future car sharing development. For this too, check the vehicle allocation confirmation, vehicle detection unit 12 and vehicle number recognition unit. This can be realized by appropriately using 13 in combination.

  In addition, the vehicle interior fullness detection device 1 of the present invention described so far has the presence or absence of damage to the vehicle from the image captured by the image capturing unit 11 when the vehicle is parked in the vehicle compartment 21 or when entering or exiting the vehicle compartment 21. And / or the degree of damage may be determined.

  As for this, the function may be added to the passenger compartment fullness detection section 12, or a function as a damage detection section may be newly provided.

  Further, instead of the image capturing unit 11, an image captured by the image capturing device (camera or smart phone) of the user may be used. In that case, the image of the damaged portion and the vehicle number is displayed on the control unit 14 of the passenger compartment fullness detection device 1 or other external system (charge adjustment system 100, monitoring system 200, or the entire parking lot not shown). May be transmitted to a parking lot control unit that controls the.

  This makes it easy to identify the time of damage, especially in the case of car sharing, so that appropriate measures can be taken.

  INDUSTRIAL APPLICABILITY The vehicle interior vacancy detection device of the present invention can reduce the installation cost and operation cost of an unmanned pay parking lot such as a coin parking, and thus has great industrial applicability.

1 Cabin Fullness Detection Device 11 Image Shooting Section 12 Cabin Fullness Detection Section 13 Car Number Recognition Section 14 Control Section 15 Display Section 16 Voice Guidance Section 17 Operation Section 18 Communication Interface Section 20 Marking Line 21 Cabin Room 22 Pole 23 Car Stop 100 Charge settlement system 200 Monitoring system

Claims (12)

An image capturing unit that captures an image of the vehicle in a passenger compartment where the vehicle is parked,
A vehicle interior fullness detection device, comprising: a vehicle interior fullness detection unit that detects a fullness state of the vehicle interior from the photographed image. The vehicle interior fullness detection device according to claim 1, further comprising a vehicle number recognition unit that recognizes a vehicle number of the vehicle from the captured image.   The vehicle according to claim 1 or 2, wherein the identity of the vehicle parked in the vehicle compartment is determined by comparing past data and current data of the vehicle compartment fullness detection unit. Room fullness detection device.   The vehicle interior fullness detection device according to claim 1 or 2, wherein the vehicle interior fullness detection unit does not determine that the vehicle is present in the vehicle interior when the vehicle movement is detected. .   The vehicle cabin fullness detection device according to claim 1 or 2, wherein when the image capturing unit has difficulty capturing an image, the information is transmitted.   The vehicle interior fullness detection unit according to claim 1 or 2, wherein the vehicle interior fullness detection unit automatically learns a parameter by using a result of another sensor to improve detection accuracy. apparatus.   The vehicle interior fullness detection device according to claim 1 or 2, wherein the detection performance of the vehicle interior fullness detection unit is improved by using it together with another sensor. The vehicle interior fullness detection device according to claim 1 or 2, wherein the vehicle interior fullness detection unit detects the presence or absence of the vehicle by recognizing a feature of the vehicle.   The vehicle interior fullness detection device according to claim 1 or 2, wherein the vehicle interior fullness detection unit detects additional information other than the presence / absence of a vehicle.   The image capturing unit or any other image capturing unit determines whether or not the vehicle is damaged and / or the degree of the damage based on an image captured when the vehicle is parked in the passenger compartment or when entering or exiting the passenger compartment. The passenger compartment fullness detection device according to claim 1 or 2.   The vehicle interior fullness detection device according to claim 2, wherein the detection result of the vehicle interior fullness detection unit is used regardless of whether or not the vehicle number recognition unit is successful.   The vehicle interior according to claim 2, wherein the vehicle interior fullness detection unit and the vehicle number recognition unit detect and notify that a previously designated vehicle is correctly parked in the vehicle interior. Full sky detection device.

Images