JP7522577B2 - Parking lot camera system, information terminal and computer program - Google Patents

Description

The present invention relates to technology for eliminating congestion in large parking lots caused by vehicles being unsure of where to park immediately after entering the parking lot.

Patent Document 1 discloses a technology to eliminate the confusion that occurs when vehicles slow down more than necessary while searching for a parking spot, causing traffic jams for following vehicles. In other words, the technology shows drivers of entering vehicles available parking spaces, monitors and guides them with cameras inside the parking lot, and allows for smooth parking.

Patent Document 2 discloses technology that allows smooth entry and exit from hourly parking lots, allows each vehicle to specify a parking space suitable for that vehicle and guides the vehicle to that parking space, provides various service information to each vehicle before and after parking, and allows smooth provision of fee discount services, etc.

As mentioned above, although there are disclosed technologies, there are still many cases in which large parking lots cause traffic congestion on surrounding roads due to vehicles attempting to enter the parking lot.
Below, we will consider the reasons for this.

For example, in the case of parking lots in large commercial facilities such as shopping malls, there are not only vehicles entering the parking lot, but also vehicles leaving the parking lot. In order to prevent accidents between vehicles leaving the parking lot and vehicles entering the parking lot, security guards visually monitor the parking lot and give instructions to the drivers of vehicles leaving the parking lot and the drivers of vehicles entering the parking lot.
In other words, automation (mechanization) that would allow for detailed instructions to be issued to prevent accidents at work sites with high traffic is expected to be difficult for technical or economic reasons (the reason why the technologies disclosed in the prior art documents have not been realized).

With reference to Figures 1 and 2, we will explain in more detail the relationship between the instructions of security guards and the vehicles that follow those instructions. The large-scale parking lot shown in both figures has multiple blocks (A, B, C, ...) in each of which multiple vehicles can be parked. Also, assume that block A is full, block B is full, and there are vacant spaces in block C.

The security guard stationed in front of Block A visually checks whether Block A, which is his/her assigned area, has vacant spaces or is full, in order to give instructions to vehicles approaching Block A. The guard then instructs vehicles (4) attempting to enter Block A to pass through if it is full, or to enter if there is vacant space. In the situation shown in Figure 1, the block is full, so vehicles (4) approaching Block A are instructed to pass through.

The driver of vehicle (4) who heard the instruction must drive slowly while keeping an eye on the movements of the next guard (in charge of Block B) and the vehicle (2) that is close to that guard. This is one of the causes of congestion in the parking lot.
If such slow driving occurs near the entrance of the parking lot, it spreads to the vehicles at the entrance, slowing down the movement of vehicles (5, 6) trying to enter the parking lot from the general road. As a result, congestion occurs on the general road starting from the entrance of the parking lot.

The security guard stationed in front of Block B also visually checks whether Block B, which is his/her area of responsibility, has vacancies or is full. Since Block B is also full, he/she will instruct vehicle (2) approaching Block B to pass through and proceed to the next block (to turn right to proceed to Block C).
The security guard stationed in front of Block C confirms that there is a vacant room and issues an instruction to park the vehicle (1) approaching Block C.

Now, the driver of vehicle (2) that has been instructed to pass in front of block B, drives while watching the movement of vehicle (1) ahead, after receiving instructions from the security guard in front of block B. This vehicle (1) attempts to enter block C, where there is a vacant room (continued in Figure 2).
However, it is often difficult to judge whether vehicle (2) can enter block C following vehicle (1) or whether the entry of vehicle (1) will cause block C to become full. Due to the uncertainty in the judgment, vehicle (2) will also start driving slowly. The slow driving of vehicle (2) will spread to the following vehicles, such as vehicle (3) and vehicle (4).

Japanese Patent Application Publication No. 09-91591 Patent No. 458750

As mentioned above, parking lot security guards are usually busy checking whether their assigned block is full or empty, while giving instructions to the drivers of vehicles trying to enter the parking lot. Meanwhile, drivers of vehicles driving through the parking lot slow down to wait for instructions from the next security guard, not knowing how far they need to go to be able to park.

An experienced security guard can check the vacancy status of his/her assigned block and keep in mind how many more vehicles can be parked, and can give instructions to the vehicle in front of him/her as well as to the vehicle behind him/her. In the case of Figure 1, while giving instructions to vehicle (1) to enter the parking lot, the guard will also signal to vehicle (2) whether it can enter or not (unlike in Figure 1, the guard must be in a position where he/she can be seen from vehicle (2)).

When drivers can see an experienced guard giving instructions, they can tell at the time they pass a certain block whether they should also pass the next block or whether they will be able to enter the parking lot. If they can do this, they can drive at a slightly higher speed instead of slowing down. This helps to ease congestion in the parking lot and also contributes to easing the chain of congestion on general roads.

However, it is difficult to train experienced security guards in a short period of time. In addition, security guards are expensive to hire, which leads to higher operating costs for parking lots.

The problem that the present invention aims to solve is to provide technology that contributes to reducing the number of security guards in large parking lots (and the burden of training them), and contributes to easing the congestion that tends to occur within large parking lots.

To solve the above-mentioned problems, we provide a first invention relating to a parking lot camera that captures and analyzes the entirety of a large parking lot, a second invention relating to a computer program that controls the parking lot camera of the first invention, a third invention relating to an information terminal that receives output from the parking lot camera of the first invention and outputs instructions to vehicles traveling in the parking lot, and a fourth invention that controls the information terminal of the third invention.

(First Invention)
The first invention relates to a parking lot camera capable of acquiring photographic data overlooking a large parking lot in which multiple vehicles can be parked in each of multiple blocks.
The parking lot camera includes a storage means for storing in advance training data for determining whether each block is occupied or not based on the photographed data, an occupancy determination algorithm, and map data of the parking lot;
An imaging means for acquiring photographic data overlooking the parking lot;
an occupancy/vacancy determination means for determining whether each block of the photographed data acquired by the imaging means is vacant or not by using the teacher data, the determination algorithm, and the map data stored in the storage means;
a transmission means for transmitting the determination result for each block determined to be full or empty by the full/empty determination means as instruction data to an information terminal disposed in each block in the parking lot;
(See Figures 3 and 4)

(Glossary)
A "large parking lot" consists of multiple blocks, and each block has multiple parking spaces (for example, 2 x 5 = 10 spaces in Figure 3) for parking vehicles. The positional relationship of each block is stored in the "map data of the parking lot."

The "imaging means" is capable of capturing an overhead image of the parking lot. It is therefore positioned higher than the ground level of the parking lot. For example, a "bird's-eye view pillar" may be used to position the parking lot camera (its lens and other imaging means) higher than the ground level of the parking lot (see Figure 3), or it may be the ceiling (or the building structure that constitutes it) of each floor except for the rooftop floor of a multi-story building-type parking lot (see Figure 13). There are cases where the entire parking lot can be captured with a single image data, and cases where the entire parking lot is captured with multiple image data, as in a building-type parking lot.

"Photographed data" refers to digital image data of the parking lot captured by the imaging means. This includes cases where all blocks of the parking lot are captured as one image data (see Figure 3), as well as cases where multiple image data are combined to capture image data of all blocks of the parking lot (see Figure 13).

The "full/vacant" state determined by the "full/vacant determination means" is not limited to a state where there are no vacant spaces in a specific block, but may also be determined to be "full" when there are only one or two vacant spaces. This is because the purpose is to ease tension in the parking lot, rather than accurately determining whether there are vacant spaces.
In order to determine whether the area is "full or empty," a large amount of pre-taken photographic data is processed using deep learning, and the determination is made based on the training data and judgment algorithm obtained from the image processing.

"Information terminal" refers not only to electronic signs installed in front of each block (see Figure 4), but also to information terminals that can be carried by security guards in charge of each block (see Figures 5 and 8).

(Action)
The storage means creates teacher data for judging whether each block is occupied or not based on photographed data captured in the past, and stores the teacher data and an algorithm for judging whether the block is occupied or not in advance.
The imaging means captures photographic data overlooking the parking lot. The vacancy determination means determines whether each block in the photographic data is vacant or not using the teacher data and a determination algorithm stored in the storage means.
The determination result for each block is transmitted by the transmitting means as instruction data to the information terminal arranged in each block.

By outputting the results of the judgment sent to the information terminal installed in each block, vehicle drivers (or security guards at each block) who can see the information terminal can know whether the block is full or empty and can quickly determine what instructions to give to vehicles. As a result, driving in the parking lot becomes smoother and contributes to easing congestion in the parking lot (and ultimately easing congestion on general roads around the parking lot).
Security guards equipped with information terminals are expected to be able to give precise instructions quickly, which is expected to contribute to smoother operation in parking lots. By relying on information terminals, security guards can be expected to demonstrate skills comparable to those of experts, even if they are not experienced.

(Variation 1 of the first invention)
In the first aspect of the present invention, the photographed data acquired by the imaging means may be transmitted to an external server (for example, a parking lot management server) that learns image processing (see Figs. 5 and 14).
In other words, if it is possible to equip the parking lot camera of the first invention with an AI function that performs machine learning image processing on the captured data, then that function may be installed (not shown).

(Glossary)
The "external server" may be a parking lot management server (see Figures 14 and 15) that collects, accumulates, and manages parking lot management data, or it may be a dedicated server that exclusively performs image processing.
"Learning image processing" means using machine learning to improve the accuracy of judging whether the sky is full or empty. "Machine learning" includes deep learning.

(Variation 2 of the first invention)
In the above-mentioned variation 1 of the first invention, it is desirable that the storage means receives and stores new teacher data newly learned by the external server executing image processing.
This is because the accuracy of determining whether a train is empty can be improved based on the new training data.

(Variation 3 of the first invention)
The first aspect of the invention is more preferably as follows.
That is, the vehicle compartment in the parking lot is provided with a distinctive part (e.g., a distinctive plate),
The occupancy determining means determines that a vehicle compartment in which a distinctive feature is recognized in the photographed data is vacant (see FIG. 12).

(Glossary)
A "characteristic part" is a part that includes a color or a light-emitting part that makes it easy to determine that the room is vacant in the photographed data acquired by the imaging means, more specifically, a color that is rarely used in vehicles, a shape that is clearly different from that of a vehicle, etc. If the vehicle interior is equipped with a flap, the flap may also play the role of a characteristic part.
Although the distinctive parts may be provided in all the vehicle compartments, it is not necessary to provide them in all the vehicle compartments. It is reasonable to provide the distinctive parts only in the areas where the determination of "full" is likely to be erroneous in the photographed data.

(Second Invention)
The second invention relates to a computer program for controlling the parking lot camera according to the first invention.
That is, a storage procedure for storing in advance in a storage means teacher data for determining whether each block is occupied or not based on the photographed data, an occupancy determination algorithm, and map data of the parking lot;
An imaging procedure for acquiring photographic data overlooking the parking lot;
an occupancy/vacancy determination step for determining whether each block of the photographed data acquired in the imaging step is occupying or not by using the teacher data, the determination algorithm, and the map data stored in the storage means;
a transmission step of transmitting the determination result for each block determined to be full or empty by the full/empty determination step as instruction data to an information terminal disposed in each block in the parking lot;
A computer program for causing the parking lot camera to execute the above.

(Variation 1 of the second invention)
The second aspect of the invention may be formed as follows.
That is, the parking lot camera is also caused to execute an external server transmission procedure for transmitting the photographed data acquired in the imaging procedure to an external server that learns image processing.

(Variation 2 of the second invention)
The second aspect of the invention may be formed as follows.
That is, in the case where the vehicle compartment in the parking lot is equipped with a special feature,
In the occupancy determination procedure, a vehicle compartment in which a distinctive feature is recognized in the photographed data may be determined to be vacant.

(Third Invention)
The third invention relates to an information terminal (an electronic signboard as shown in Figures 3 and 4, etc., and a security guard terminal as shown in Figures 8 and 7) that is installed near the vehicle entrance of each of a number of blocks in a large parking lot in which a number of vehicles can be parked in each of the blocks.
That is, an installation location means for associating the location information of each block in the parking lot with the location information of each block;
An instruction data receiving means for receiving instruction data based on whether each block in the parking lot is full or empty;
a display means for outputting output data based on the position information associated by the installation position means and the instruction data received by the instruction data receiving means;
(See FIG. 4.)

(Glossary)
The "installation location means" may be a device using short-range wireless communication, as described below, or may be a so-called GPS function.

(Action)
The location information storing means stores information on the location of each block in the parking lot, and the location information storing means stores information on the location of each block in the parking lot.
The display means outputs the output data based on the position information associated by the installation position means and the instruction data received by the instruction data receiving means.

When the information terminal is an electronic signboard, the driver of a vehicle traveling in the parking lot can visually check the display means on which the output data is displayed, thereby contributing to smooth driving in the parking lot.
If the information terminal is a security guard terminal, the security guard can visually check the display means on which the output data is displayed (without having to visually check the occupancy status of his/her assigned block in the parking lot), which helps to quickly issue accurate instructions (even if he/she is not an experienced security guard) to drivers of vehicles traveling in the parking lot.
By combining electronic signs with security guards, the number of security guards can be reduced compared to when all staff are security guards, which helps to reduce labor costs.

(Variation 1 of the third invention)
The installation location means in the third invention may acquire location information (location information of each block) by two-way communication with short-range wireless communication means installed in each block in the parking lot.

A "short-distance wireless communication means" is a device that starts two-way wireless communication when it comes within about 10 to 20 meters of the device, and is called a wireless PAN (Personal Area Network). A typical example of such a device is Bluetooth (a registered trademark of The Bluetooth SIG Inc.).

(Action)
Since location information (location information for each block) is obtained through two-way communication with short-range wireless communication means installed in each block in the parking lot, when multiple information terminals are provided, there is no need to use dedicated terminals associated with location information.

(Variation 2 of the third invention)
It is more preferable that the display means in the third aspect of the invention is formed so that the height from the ground surface of the block on which it is installed is adjustable (see FIG. 11).

(Action)
If the height of the display means is adjusted to be higher, it will be more easily visible to drivers of vehicles located farther away from the location where the information terminal is installed (compared to when the display means is not adjusted higher).
If visibility from a distance becomes possible, the driver of the vehicle will have more options and information to decide how to proceed next, which will lead to smoother driving, and as a result, it may contribute to reducing congestion in parking lots.

(Variation 3 of the third invention)
It is more preferable that the instruction data receiving means in the third invention is configured to receive instruction data from an image processing server which acquires photographic data of an overhead view of the parking lot and performs image processing to determine whether each block is occupied or not (see Figures 4, 9 and 16).

If the parking lot is capable of knowing whether each individual space is occupied or not, instruction data can be received from a management server that knows whether each space in the parking lot is occupied or not.
Here, in parking lots where it is not possible to know whether each parking space is occupied or not, the system substitutes for this by acquiring photographic data that shows an overhead view of the parking space. This eliminates the need for equipment to determine whether each parking space is occupied or not.

(Fourth Invention)
A fourth aspect of the present invention relates to a computer program for controlling the information terminal according to the third aspect of the present invention.
That is, an installation location procedure for associating the location information of each block in the parking lot with the location information of each block in the parking lot;
an instruction data receiving step of receiving instruction data based on whether each block in the parking lot is full or empty;
a display step of outputting data based on the position information associated in the installation position step and the instruction data received in the instruction data receiving step;
The computer program is adapted to cause the information terminal to execute the above.

(Variation 1 of the fourth invention)
In the installation location step according to the fourth aspect of the present invention, the location information may be obtained by two-way communication with short-range wireless communication means installed in each block in the parking lot.

(Variation 2 of the fourth invention)
In the instruction data receiving procedure in the fourth aspect of the present invention, the instruction data may be received from an image processing server that acquires photographic data of an overhead view of the parking lot and processes the images to determine whether each block is occupied or not.

The computer programs according to the second and fourth aspects of the present invention can also be transmitted from a computer storing the programs to other terminal means via a communication line.
The computer programs according to the second and fourth aspects of the invention can also be provided by storing them in a recording medium. Here, a "recording medium" is a medium capable of carrying a program that does not occupy space by itself. Examples of such a medium include a hard disk, a CD-R, a DVD-R, and a flash memory.

According to the first invention, it is possible to provide a parking lot camera that contributes to easing congestion within large parking lots that tends to occur inside the parking lots.
According to the second invention, it is possible to provide a control program for a parking lot camera that contributes to easing congestion within a large parking lot that tends to occur inside the parking lot.
According to the third aspect of the invention, it is possible to provide an information terminal that contributes to easing congestion within large parking lots that tends to occur inside the parking lots.
According to the fourth aspect of the present invention, it is possible to provide a control program for an information terminal that contributes to easing congestion within a large parking lot that tends to occur inside the parking lot.

FIG. 1 is a conceptual diagram for explaining problems in a conventional large-scale parking lot. FIG. 1 is a conceptual diagram for explaining problems in a conventional large-scale parking lot. FIG. 1 is a perspective view showing an embodiment of the present invention. 1 is a block diagram showing an embodiment of the present invention focusing on a parking lot camera and an electronic signboard. This is a conceptual diagram showing image processing AI that creates training data for determining whether a seat is full or empty. 13 is a flowchart for creating training data for distinguishing between full and empty. 13 is a flowchart showing a procedure from photographing a parking lot to creating and transmitting instruction data. FIG. 11 is a perspective view showing a case where an electronic signboard and a security guard terminal are used in combination in an embodiment of the present invention. FIG. 1 is a block diagram focusing on a security guard terminal. FIG. 2 is a perspective view showing various dimensions of the electronic signboard. FIG. 11 is a perspective view showing another embodiment of the digital signboard. This is a perspective view showing that a special feature plate is provided for each parking space in a parking lot. This is a side view showing a large-scale parking lot as a three-story parking building. 11 is a block diagram showing a case where instruction data is common to each terminal in information processing within the parking lot camera. FIG. 11 is a block diagram showing a case where instruction data is dedicated to each terminal in information processing within the parking lot camera. FIG. FIG. 11 is a block diagram showing an embodiment in which instruction data for an electronic signboard is transmitted from a parking lot management server.

The present invention will be described below based on the drawings and embodiments. The drawings used here are Figures 3 to 16.

(Figure 3)
Figure 3 shows a perspective view of (a part of) a large parking lot with multiple blocks, each with multiple parking spaces, and the relationship with the public roads through which vehicles can enter the parking lot.
The parking lot camera, which is a digital camera equipped with functions such as sending and receiving data, is fixed to an overhead pillar to ensure sufficient height for shooting so that it can capture all the blocks of the large parking lot, which consists of many blocks on a vast site, in a single image.

The large-scale parking lots are intended to accommodate the vehicles of people visiting large commercial facilities, entertainment facilities, and the like.
The parking lot camera captures images and captures image data every minute during the hours that the commercial facility or the like is open.
Furthermore, if the parking lot camera is to take pictures at night, it will be necessary to use bright lenses and highly sensitive optical elements, and to provide sufficient lighting in the parking lot.

In Fig. 3, Blocks A, B, and C are shown as parts of a large parking lot. Electronic signs are installed near the entrances of each block. As will be described in detail in Fig. 4, the electronic signs display whether the block in which they are installed is full or whether parking is possible, and if it is full, they also display the direction (passing instruction display) in which to move past the electronic sign to park.
The sample vehicle shown in FIG. 3 will be explained after the explanation of FIG.

(Figure 4)
Fig. 4 shows the detailed configuration of the parking lot camera and electronic signboard shown in Fig. 3, and a part of the parking lot (Block B and the partition paint for separating Block B from other parts). In the following figures, wireless communication is indicated by a dashed double-dashed arrow, and wired communication is indicated by a solid arrow.

The parking lot camera is equipped with an imaging means for capturing images of the entire parking lot (every minute during business hours), an occupancy determination means for determining whether each block is full or has vacancies based on the captured image data, and a transmission means for transmitting instruction data resulting from the determination by the occupancy determination means to the electronic signboard. It is also equipped with a storage means for storing in advance the determination algorithm used by the occupancy determination means, teacher data, and map data of the parking lot for determining whether each block is full or vacant.

The instruction data transmitted from the transmitting means of the parking lot camera is received by the receiving means of the electronic signboard.
The electronic signboard has a built-in short-range wireless communication chip, and receives block data (B) through two-way communication with a short-range wireless communication chip fixed, for example, by burying it in the road surface of the block, and knows where the electronic signboard has been installed.

The instruction data received by the receiving means includes instruction content for each block. From the instruction data, the control means extracts the instruction content of the block data (B) and transmits it to the display means as output data.

The display means are a "FULL" lamp indicating that the parking lot is full, a "Vacant" lamp indicating that there is a vacant space, and a passing indicator (arrow lamp) indicating the direction to pass through the block and the direction to head in. The "FULL" lamp or the "Vacant" lamp are selectively lit. In addition, the passing indicator is also lit when the "FULL" lamp is lit. This diagram shows that the passing indicator and the "FULL" lamp are lit.

The photographed data is updated periodically (this can be changed in the settings, but in this embodiment, every minute), so the instruction data transmitted along with it is also determined by the occupancy determination means at any time and transmitted every minute. Therefore, the output data output to the electronic signboard is also revised every minute. Note that the interval between photographs can be set to less than one minute (60 seconds), and the photographed data can be a video.

The explanation will be given on the assumption that the vacant/empty sign and the passing instruction sign in the electronic sign shown in Fig. 4 can be displayed on both sides of the sign, but the display (lighting) on the side facing the block side may be omitted. This is for the purpose of saving power, since this electronic sign runs on battery power, as shown in Fig. 10.
Furthermore, since the display on the side facing the block is not very important (there is little need to see the electronic sign from within the block), it is also possible to provide an electronic sign that does not have a display device facing the block.

Figure 3 explains how an electronic signboard and parking lot camera with the functions shown in Figure 4 act on a sample vehicle that enters the large parking lot of this embodiment from a public road.

(Sample vehicle)
The driver of the sample vehicle that passed through block A shown in Figure 3 can see the electronic sign installed in front of block B. The electronic sign has a passing instruction display and a "full" character lamp lit up. Therefore, the moment the driver sees the electronic sign installed in front of block B, the driver can decide to pass through block B as well. From the driver of the vehicle following the sample vehicle (not shown in this figure), it can be seen that the sample vehicle is also trying to pass through block B, and if the driver sees the electronic sign installed in front of block B, the driver can decide that he or she should also pass through block B.

The photographic data captured by the imaging means of the parking lot camera is not only used to create instruction data, but is also sent to the image processing AI server to improve the accuracy of the vacancy determination means. The transmitted photographic data is received by the receiving means, undergoes image learning, and is then transmitted from the transmitting means to the parking lot camera. It is then stored as new teacher data in the storage means.

(Figure 5)
FIG. 5 conceptually illustrates details of the image processing AI server shown in FIG.
In the image processing AI server, the already captured photographed data is classified into an image file of vacant spaces and an image file of full spaces, and stored in the storage means. For convenience of illustration, an image file of vacant spaces and an image file of full spaces are shown for the image data of three blocks, but in reality, a single image data of the entire parking lot is divided into each block and processed into each block photo data (this processing process can be automated relatively easily). Then, a person in charge of classification classifies each block photo data into an image of vacant spaces or an image of full spaces (so-called "annotation").

Based on the results of classifying each block photo data into an image with vacant rooms or an image with full parking spaces, training data is separately created to distinguish whether each block photo data before classification is vacant or full. The created training data and the judgment algorithm are also stored in the storage means.

The photographic data received by the receiving means is calculated using the judgment algorithm and teacher data of the storage means to create new teacher data (illustrated as "new teacher data"). The new teacher data is then stored in the storage means and also transmitted to the parking lot camera via the transmission means. After that, the teacher data is replaced with the new teacher data in the parking lot camera.

In FIG. 14, the image processing AI is described as being provided within the parking lot management server, but it may also be a server dedicated to image processing.
The image processing AI can be built into the parking lot camera as in the embodiment. When built into the parking lot camera as shown in Fig. 4, the receiving means and the transmitting means in Fig. 14 are not required.

(Figure 6)
6 shows a flowchart for creating training data for distinguishing between full and empty parking spaces. This flowchart includes a step of distinguishing between full and empty parking spaces and a step of creating training data, and is not a flowchart showing a fully automated procedure.

First, the parking lot camera's imaging means takes a wide-angle bird's-eye view of the parking lot so that the entire parking lot can be captured on one image, and acquires photo data (S1). The acquired photo data is divided into blocks, and each block's photo data is generated (S2). Then, each block's photo data is stored, distinguishing whether it is full or has vacant spaces (S3).

Teacher data that can distinguish between full and empty is created for each block (S4).
We will not go into detail about how the training data is created, but we decided to create training data for each block because many parameters of the photo data vary, such as the position of the sun in each season, the time of day, and the color of the parked vehicles, and the effects of these many parameters may differ for each block.

(Figure 7)
Fig. 7 shows a flowchart showing the procedure for creating and sending instruction data using the teacher data for photo data of a parking lot after creating the teacher data shown in Fig. 6. All steps such as photography, image processing, and instruction data creation can be automated.

The parking lot camera captures a wide-angle image of the parking lot from above so that the entire parking lot can be captured in one shot, and acquires photographic data (S11). The acquired photographic data is divided into blocks, and each block of photographic data is generated (S12).
Next, using the teacher data and judgment algorithms that have already been created, it is judged whether each block of photo data is full or empty (S13). Then, based on the result of the vacancy/empty judgment, instruction data for each block is created and the instruction data is transmitted (S14).

When one minute has passed since the acquisition of the photo data (S11) ("Yes" in S15), it is determined whether the parking lot has closed for business (S16), and if it is within business hours, an aerial photograph of the parking lot is taken again to acquire photo data (S11).

The purpose of the present invention is to facilitate the movement of vehicles within a large-scale parking lot. Therefore, although it has been described that in S16, it is determined whether or not the parking lot's closing time has arrived, it may be determined that the time has come when congestion will no longer occur within the parking lot, rather than the closing time.
Also, the process may end when it is determined that the parking lot is not crowded using a different algorithm, such as taking an overhead shot of the parking lot to determine how crowded it is.

The result of the vacant/empty determination made in S13 may be judged as to whether or not the result is correct, and the teacher data may be revised (new teacher data may be created) (S17).
In addition, in Figures 4 and 5, the process is described as including a step of creating new training data.

(Figure 8)
8 shows that instead of an electronic signboard, a security guard is standing in front of Block B, and the security guard is holding a security guard terminal equipped with the same functions as the electronic signboard. The security guard terminal is shown in FIG. 9.

(Figure 9)
In Fig. 9, the electronic signboard shown in Fig. 8 has been replaced with a security guard terminal, but the functions of the security guard terminal are the same as those of the electronic signboard. For convenience of illustration, the terminal is shown larger than the security guard, but in reality it is small enough to fit in the palm of the guard's hand.

The security guard terminal shown in FIG. 9 can also be realized by installing a specific application on a smartphone equipped with a short-range wireless communication chip.
The built-in short-range wireless communication chip performs two-way communication with the short-range wireless communication chip of the block, extracts output data from the received instruction data, and outputs it to the display means.

(Figure 10)
FIG. 10 shows the external appearance of the digital signage in detail.
This electronic sign is equipped with a wheeled base that is 600 mm wide and 300 mm deep, and is 1200 mm high. The vacant/empty sign and passing instruction sign, both of which have a diameter of 400 mm, are positioned high up (800 mm above ground level) to make them as easy to see as possible for drivers and security guards from a distance. Note that the dimensions of the electronic sign are not limited to the values in this embodiment.

The full/vacant indicator and the passing indicator are turned on and off by lighting the built-in LEDs. This diagram shows the passing indicator and the "full" indicator turned off, and the "vacant" indicator turned on. The electrical energy required to light them is supplied by a built-in battery.

This digital sign can be moved using wheels attached to its base. The wheels can also be locked when the sign is positioned in front of a designated block. Since it is equipped with a short-range wireless communication chip, it can communicate two-way with short-range wireless communication chips buried in designated locations in each block, allowing the sign to determine where it is.

(Figure 11)
Fig. 11 shows a variation of the electronic signboard. The electronic signboard shown here has a double-structure pedestal with an X-link between the two pedestals. By adjusting the X-link, the height of the vacant/empty indicator and the passing instruction indicator on the electronic signboard can be adjusted.

By increasing the height of the vacant/empty sign and the passing instruction sign, visibility is improved for drivers of vehicles traveling at a distance, especially in locations closer to the parking lot entrance than the electronic sign. This allows the driver to obtain information to decide on their course of action at an earlier stage, which helps to improve driving speeds within the parking lot. This therefore helps to alleviate congestion within the parking lot.

(Figure 12)
Fig. 12 shows an embodiment in which a "feature plate" is fixed to the road surface of each vehicle compartment. This feature plate has a color and shape that makes it easy to distinguish when a vehicle compartment is vacant in the image data captured by the parking lot camera. This contributes to improving the accuracy of judging whether the vehicle compartment is vacant or not in the image data.
When a vehicle is parked in the vehicle compartment, the vehicle compartment feature plate is positioned so that it does not appear in the captured data even if the parking lot camera takes an image.

When parking spaces are equipped with flaps, the flaps serve as distinctive features, but this large-scale parking lot does not have flaps on the parking spaces because it is easy or unnecessary to pay parking fees.

(Figure 13)
FIG. 13 shows what a parking lot camera would look like if the large parking lot were a parking building.
In this figure, the parking lot building is two stories tall with a rooftop. A parking lot camera (1) is installed on the ceiling of the first floor, a parking lot camera (2) is also installed on the ceiling of the second floor, and a parking lot camera (3) is installed on an overhead pillar installed on the rooftop (RF).

When using multiple cameras to obtain photographic data for each block, it is reasonable to obtain training data for each camera. This is because each block has its own characteristics, and the determination of whether an area is empty or not depends on those characteristics.

(Figure 14)
FIG. 14 shows a block diagram in which a parking lot management server is provided in a location physically separate from the parking lot cameras, and the parking lot management server includes an image processing AI.
The parking lot camera transmits instruction data to each information terminal, which are assumed to be an electronic signboard (A), a security guard terminal (B), and an electronic signboard (C) (the number of information terminals corresponds to the number of blocks).

The instruction data sent from the parking lot camera to each information terminal is common to all terminals. Each information terminal extracts only the instruction data corresponding to its own block position obtained by the short-range wireless communication chip, and uses it as output data.

(Figure 15)
In FIG. 15 as well, a parking lot management server is provided in a location physically separate from the parking lot cameras, and the parking lot management server includes an image processing AI.
The difference from FIG. 14 is that the instruction data created by the parking lot vacancy determination means in the parking lot camera is created for each information terminal.

In the case of the parking lot vacancy determination means in the parking lot camera shown in Figure 15, there is no need to separate the instruction data in the information terminal. This has the advantage that it is possible to omit the short-range wireless communication chip described as being provided in the information terminal and each block.

However, each information terminal must be individually identified in advance and configured to correspond to each block before it can be used. In other words, each information terminal is given the name of a block, and it cannot be used in any block other than the one it is named for.

(Figure 16)
FIG. 16 shows an embodiment in which the intelligent functions in the parking lot camera described so far are transferred to the parking lot management server.

The "parking lot photography camera" shown in FIG. 16 is similar to the parking lot camera in that it is installed on the overhead pillar shown in FIG. 3 and captures an entire large parking lot in one image.
The functions that the parking lot camera internally processes and judges are assumed by the parking lot management server. In other words, the parking lot management server is equipped with a receiving means for receiving the photographed data photographed by the parking lot photographing camera, a storage means for storing the teacher data, the judgment algorithm, the map data of the parking lot, etc. in advance, an occupancy/vacancy judgment means for judging the occupancy/vacancy of each block of the parking lot using the photographed data, the teacher data, the judgment algorithm, and the map data of the parking lot, and a transmission means for transmitting the instruction data judged by the occupancy/vacancy judgment means to an information terminal.

The above-described embodiment contributes to mitigating congestion within large parking lots that tends to occur inside the parking lots.
Furthermore, if the security guard terminal shown in FIG. 8 and FIG. 9 is used in combination with an electronic signboard, a shortage of security guards can be addressed.

Even if electronic signs are not used and only security guard terminals are used, the chances of issuing appropriate instructions to vehicles attempting to park can be increased even if the security guard does not have much experience.

A parking lot using the digital signage according to the embodiment described above can quickly and accurately issue instructions to vehicles trying to park, which contributes to reducing congestion in the parking lot and reducing congestion on general roads connected to the parking lot that are caused by the parking lot.

The present invention can be used in the parking lot operation and management industry, the manufacturing industry of parking lot cameras and information terminals, the data service industry that handles data related to parking lot cameras and information terminals, and the image processing service industry of image data captured by parking lot cameras.

Claims (15)

A parking lot camera system comprising: a parking lot camera capable of acquiring photographic data overlooking a large parking lot in which a plurality of vehicles can be parked in each of a plurality of blocks ; and an information terminal disposed in each block in the parking lot for receiving and outputting data transmitted from the parking lot camera;
The parking lot camera is
A storage means for storing in advance training data for determining whether each block is occupied or not based on the photographed data, an occupancy determination algorithm, and map data of the parking lot;
An imaging means for acquiring photographic data overlooking the parking lot;
an occupancy/vacancy determination means for determining whether each block of the photographed data acquired by the imaging means is occupying or not by using the teacher data, the determination algorithm, and the map data stored in the storage means;
a transmission means for transmitting the determination result for each block determined to be full or empty by the vacancy determination means as instruction data to an information terminal disposed in each block in the parking lot;
The following measures will be taken :
The information terminal includes:
an installation location means for associating the location information of each block in the parking lot with the location information of each block;
an instruction data receiving means for receiving instruction data based on the occupancy status of each block in the parking lot from the parking lot camera;
a display means for outputting output data based on the position information associated by the installation position means and the instruction data received by the instruction data receiving means;
A parking lot camera system comprising : 2. The parking lot camera system according to claim 1, wherein the photographed data acquired by the imaging means is transmitted to an external server that learns image processing. 3. The parking lot camera system according to claim 2, wherein the storage means receives and stores new teacher data newly learned by the external server executing image processing. The parking lot has a special feature in the parking space.
4. A parking lot camera system according to claim 1, wherein the occupancy determining means determines that a vehicle compartment in which a distinctive feature is recognized in the photographed data is vacant. The installation location means acquires location information by two-way communication with short-range wireless communication means installed in each block in the parking lot.
The parking lot camera system of claim 1 . 2. The parking lot camera system according to claim 1 , wherein the height of the display means in the information terminal from the ground surface of the block in which it is installed is adjustable. The instruction data receiving means receives instruction data from an image processing server that acquires photographic data of a bird's-eye view of the parking lot and processes the images to determine whether each block is full or empty.
The parking lot camera system of claim 1 . An image processing server that revises the teacher data,
The image processing server includes an image data receiving means for receiving image data acquired by the parking lot camera;
A storage means for storing the teacher data in advance;
An image learning means for creating new teacher data, which is new teacher data, using the image data received by the image data receiving means and the teacher data stored in the storage means;
A transmission means for transmitting new teacher data created by the image learning means to the parking lot camera.
The parking lot camera system of claim 1 . A computer program for controlling a parking lot camera system including parking lot cameras capable of acquiring bird's-eye view photographic data of a large parking lot in which a plurality of vehicles can be parked in each of a plurality of blocks, and information terminals disposed in each block in the parking lot for receiving and outputting data transmitted from the parking lot cameras ,
a storage step of storing in advance in a storage means teacher data for determining whether each block is occupied or not based on the photographed data, an occupancy determination algorithm, and map data of the parking lot;
An imaging procedure for acquiring photographic data overlooking the parking lot;
an occupancy/vacancy determination step for determining whether each block of the photographed data acquired in the imaging step is occupying or not by using the teacher data, the determination algorithm, and the map data stored in the storage means;
a transmission step of transmitting the determination result for each block determined to be full or empty by the full/empty determination step as instruction data to an information terminal disposed in each block in the parking lot;
The parking lot camera executes the above.
An installation location procedure for associating the location information of each block in the parking lot with the location information of each block;
an instruction data receiving step of receiving instruction data based on the occupancy status of each block in the parking lot from the parking lot camera;
a display step of outputting data based on the position information associated in the installation position step and the instruction data received in the instruction data receiving step;
The above-mentioned information terminal is to execute the above.
Computer program. In the installation location procedure, location information is acquired by two-way communication with short-range wireless communication means installed in each block in the parking lot.
10. A computer program according to claim 9 . In the above-mentioned instruction data receiving procedure, the instruction data is received from an image processing server that acquires photographic data of a bird's-eye view of the parking lot and processes the images to determine whether each block is full or empty.
A computer program according to claim 9 or 10 . A parking lot camera system includes a parking lot camera capable of acquiring photographic data overlooking a large parking lot in which a plurality of vehicles can be parked in each of a plurality of blocks , and an information terminal that is disposed near the vehicle entrance of each block in the parking lot and receives and outputs data transmitted from the parking lot camera ,
The parking lot camera is
A storage means for storing in advance training data for determining whether each block is occupied or not based on the photographed data, an occupancy determination algorithm, and map data of the parking lot;
An imaging means for acquiring photographic data overlooking the parking lot;
an occupancy/vacancy determination means for determining whether each block of the photographed data acquired by the imaging means is occupying or not by using the teacher data, the determination algorithm, and the map data stored in the storage means;
a transmission means for transmitting the determination result for each block determined to be full or empty by the full/empty determination means as instruction data to an information terminal disposed in each block in the parking lot,
The information terminal includes:
an installation location means for associating the location information of each block in the parking lot with the location information of each block;
an instruction data receiving means for receiving instruction data from the parking lot camera as a result of a judgment made by the parking lot camera based on the occupancy or vacancy of each block in the parking lot;
a display means for outputting output data based on the position information associated by the installation position means and the instruction data received by the instruction data receiving means;
An information terminal equipped with 13. The information terminal according to claim 12, wherein the installation location means acquires the location information by two-way communication with short-range wireless communication means installed in each block in the parking lot. The display means is adjustable in height from the ground surface of the block where it is installed.
13. An information terminal according to claim 12 . The information terminal according to claim 12 or 13, wherein the instruction data receiving means receives instruction data from an image processing server that acquires photographic data of an overhead view of the parking lot and performs image processing to determine whether each block is occupied or not.

Images