JP7302511B2 - Information processing equipment - Google Patents

Description

The present invention relates to an information processing device, an information processing method, and an information processing program for processing information acquired by an infrastructure sensor and transmitting the information to a vehicle.

In recent years, developments related to automatic driving of automobiles have been advanced. In particular, the development of self-driving vehicles, in which an on-board system automatically accelerates, steers, and brakes the vehicle, and the driver manually drives the vehicle only when requested by the system, is being actively developed. Such self-driving vehicles control the running of the vehicle based on information received via road-to-vehicle communication or vehicle-to-vehicle communication, or information obtained using sensors mounted on the vehicle.

One of the information transmitted and received by road-to-vehicle communication is information acquired by an infrastructure sensor. Infrastructure sensors installed along roads detect not only vehicles traveling within the sensor's detectable range, but also objects such as obstacles and pedestrians on the road, and provide the detected information to vehicles. . Information provided by infrastructure sensors includes a wide range of information that is difficult to obtain with sensors mounted on vehicles, so it can be useful information for self-driving vehicles.

By the way, if there is any object within the range detectable by the infrastructure sensor, the area behind it becomes a blind spot for the sensor. Even if there is another object in such a blind area, the infra-sensor cannot detect it. In such a case, the infrastructure sensor preferably provides the vehicle with information on the blind area that the sensor cannot detect, in addition to the information on the object detected by the sensor, thereby alerting the vehicle to the blind area.

For example, Patent Literature 1 discloses calculating a blind spot area caused by an object such as a vehicle recognized by an infrastructure sensor, and transmitting the object information and the blind spot information to a vehicle traveling near the infrastructure sensor. According to the invention described in Patent Literature 1, the vehicle enables safe driving by performing driving control of the own vehicle based on information obtained from the infrastructure sensor and information obtained from the in-vehicle sensor.

JP 2018-195289 A

Here, the blind spot area that cannot be detected by the infrastructure sensor is actually a three-dimensional area. Therefore, it is conceivable that the infrastructure sensor transmits blind spot information indicating a three-dimensional blind spot area to the vehicle. However, when a vehicle calculates its own driving plan, it often maps the surrounding environment information on a two-dimensional map and processes it, and converts the received three-dimensional blind spot information into planar two-dimensional blind spot information. This increases the processing load on the vehicle. Moreover, when the infrastructure sensor transmits the three-dimensional blind spot information to the vehicle, there is a possibility that the amount of communication between the infrastructure sensor and the vehicle increases. Therefore, it is desirable that the infrastructure sensor converts the three-dimensional blind spot information into two-dimensional blind spot information and transmits the converted information to the vehicle.

Here, the inventor found out the following problems.
Infrastructure sensors can detect objects that are taller than the vehicle, even if they are in the blind spot area behind the vehicle. However, when the infrastructure sensor transmits to the vehicle two-dimensional blind spot information indicating a blind spot area caused by the vehicle and information about an object existing in the blind spot area, the vehicle detects an object in the blind spot area that the infrastructure sensor cannot detect. You will receive conflicting information that there is a As a result, there is a risk that vehicle blind spot information and information about detected objects will not be used when an autonomous vehicle calculates a travel plan.

Therefore, in the present invention, in addition to information indicating the two-dimensional blind spot area, information indicating the height of the blind spot area is transmitted from the infrastructure sensor to the vehicle as blind spot information, so that the sensor detects an object existing in the blind spot area. To provide a vehicle with object information and blind spot information that are not inconsistent with each other even when the vehicle is in a closed position.

An information processing device (10, 20) according to an aspect of the present disclosure includes an image acquisition unit (102) that acquires an image detected by an infrasensor, an object detection unit (103) that detects an object included in the image, a blind spot calculation unit (104) for calculating a planar area and a height of each divided blind spot area obtained by dividing the three-dimensional blind spot area caused by the subject into a plurality of blind spot areas; and the planar area and the height of the divided blind spot area. and transmission units (105, 106) for transmitting blind spot information including

It should be noted that the numbers in parentheses attached to the constituent elements of the invention described in the claims and this section indicate the corresponding relationship between the present invention and the embodiments described later, and are not intended to limit the present invention. do not have.

With the configuration as described above, the information processing apparatus can transmit blind spot information including the planar area and height of the blind spot area. As a result, even if the vehicle that communicates with the infrastructure sensor receives both the blind spot information and the information about the object existing in the blind spot area indicated by the blind spot information, the vehicle does not recognize that these pieces of information contradict each other and does not receive the blind spot information. And it becomes possible to calculate a travel plan using the information of the object.

FIG. 2 is a diagram for explaining the infrasensor systems of the first and second embodiments; FIG. 2 is a block diagram for explaining the configuration of the infrastructure sensor device according to the first and second embodiments; FIG. 4 is a diagram for explaining messages generated by message generation units according to Embodiments 1 to 3; FIG. 4 is a diagram for explaining a subject detected by the sensor units of Embodiments 1 to 3 and a blind area caused by the subject; FIG. 5 is a diagram for explaining planar regions and heights of blind spot regions and divided blind spot regions according to the first embodiment; FIG. 5 is a diagram for explaining the planar area and height of the divided blind area according to the first embodiment; FIG. 4 is a diagram for explaining blind spot information and subject information according to the first embodiment; FIG. 4 is a diagram for explaining the operation of the information processing apparatuses according to the first to third embodiments; FIG. 5 is a diagram for explaining the operation of the blind spot calculation unit of the first embodiment; The figure explaining the plane area|region and height of the division|segmentation blind spot area|region of Embodiment 2. The figure explaining the plane area|region and height of the division|segmentation blind spot area|region of Embodiment 2. FIG. 10 is a diagram for explaining an infrastructure sensor system according to a third embodiment; Block diagram for explaining the configuration of the server device and the infrastructure sensor device of the third embodiment

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described with reference to the drawings.

It should be noted that the present invention means the invention described in the scope of claims or the section of Means for Solving the Problems, and is not limited to the following embodiments. In addition, at least the words and phrases in angle brackets mean the words and phrases described in the claims or the means for solving the problems section, and are not limited to the following embodiments.

The configurations and methods described in the dependent claims are arbitrary configurations and methods in the invention described in the independent claims. The configurations and methods of the embodiments corresponding to the configurations and methods described in the dependent claims, and the configurations and methods described only in the embodiments without being described in the claims, are optional configurations and methods in the present invention. The configuration and method described in the embodiment when the description of the claims is broader than the description of the embodiment are also arbitrary configurations and methods in the present invention in the sense that they are examples of the configuration and method of the present invention. . In either case, the essential features and methods of the invention are described in the independent claims.

The effects described in the embodiments are the effects when having the configuration of the embodiment as an example of the present invention, and are not necessarily the effects of the present invention.

When there are multiple embodiments, the configuration disclosed in each embodiment is not limited to each embodiment, but can be combined across the embodiments. For example, a configuration disclosed in one embodiment may be combined with another embodiment. Further, the configurations disclosed in each of the plurality of embodiments may be collected and combined.

The problems described in the Problems to be Solved by the Invention are not known problems, but were independently discovered by the inventors, and are facts that affirm the inventive step of the invention together with the structure and method of the present invention.

(Embodiment 1)
1. Infrastructure sensor system 1
FIG. 1 is a diagram for explaining an infrastructure sensor system 1. As shown in FIG. The infrastructure sensor system 1 includes an infrastructure sensor device 10 (corresponding to an “information processing device”) and a communication device 11 .

The infrastructure sensor device 10 is a device installed on a road or a building around the road. road signs, buildings, etc.).

The communication device 11 is a communication device mounted on a mobile object to receive messages transmitted by the infrastructure sensor device 10 . FIG. 1 shows an in-vehicle communication device 11 mounted in a vehicle. A vehicle equipped with the communication device 11 utilizes the message received by the communication device 11 from the infrastructure sensor device 10 and reflects the message in the travel plan of the vehicle. Note that the communication device 11 is not limited to an in-vehicle communication device, and may be, for example, a communication device mounted on a mobile terminal carried by a pedestrian.

2. Configuration of Infrastructure Sensor Device 10 The configuration of the infrastructure sensor device 10 will be described with reference to FIG. The infrastructure sensor device 10 includes a sensor section 101 , an image acquisition section 102 , a subject detection section 103 , a blind spot calculation section 104 , a message generation section 105 and a transmission section 106 .

A sensor unit (corresponding to an “infrastructure sensor”) 101 is a sensor that detects an object existing on a road or around the road. The sensor unit 101 is, for example, a camera for capturing a moving image or still image of an object, or a distance sensor such as LiDAR. The sensor unit 101 is a sensor that detects information necessary to acquire an image, which will be described later, but may also detect the speed and moving direction of an object.

The image acquisition unit 102 acquires the “image” of the object detected by the sensor unit 101 . If the information detected by the sensor unit 101 is not video information such as a moving image or a still image, the image acquisition unit 102 converts the information detected by the sensor unit 101 to generate an image. good.

Here, the "video" may be a still image as well as a moving image composed of one or more images. In addition to images captured through a camera lens, images obtained by converting three-dimensional data obtained by scanning with a distance sensor such as LiDAR (Light Detection And Ranging) into two-dimensional data may be used.

A subject detection unit 103 detects a subject included in the image acquired by the image acquisition unit 102 . When the subject detection unit 103 detects a subject included in the video, it outputs subject information about the detected subject to the blind spot detection unit 104 . The subject information includes, for example, subject identification information, subject position coordinates, occupied area size, movement direction and speed, and subject type determined based on subject shape (e.g., vehicle, pedestrian, building, etc.). included. The moving direction and speed of the subject may be information detected by the sensor unit 101, or may be information acquired by analyzing the video.

The blind spot calculation unit 104 calculates a three-dimensional blind spot area caused by the subject based on the subject information output from the subject detection unit 103 . The blind spot calculation unit 104 further “calculates” the planar area and “height” of each divided blind spot area obtained by dividing the calculated three-dimensional blind spot area into a plurality of blind spot areas. The method by which the blind spot calculator 104 calculates the planar area and height of the divided blind spot area will be described later. It should be noted that the blind spot calculation unit 104 only needs to be able to calculate the plane area and the height of the divided blind area, and does not need to calculate the divided blind area itself obtained by dividing the three-dimensional blind area.

Here, the "height" may be a value representing the distance in the vertical direction with respect to the planar area of the divided blind area, for example, the maximum height, minimum height, average height, and Including combinations of these.
"Calculating" means obtaining the planar area and the height by performing calculations, as well as setting either the planar area or the height to a preset value and determining the other of the planar area or the height. including. It also includes setting the optimum plane area and height using a predetermined table.

The blind spot calculation unit 104 outputs the subject information and the blind spot information including the planar areas and heights of the divided blind spot areas to the message generating unit 105 .

A message generation unit 105 generates a message including subject information and blind spot information. FIG. 3 shows an example of a message generated by the message generator 105. As shown in FIG. As shown in FIG. 3, the message includes subject information and blind spot information. In the example shown in FIG. 3, the subject information includes the number (N) of subjects included in the video, the ID and type information of each of the N subjects, and the coordinate information of the subject. In addition, the blind spot information includes the number (M) of split blind spot areas, the ID of the subject that is the cause of each of the M split blind spot areas, the coordinate information of the split blind spot areas, and the height information. In the example shown in FIG. 3, the subject IDs included in blind spot information #1 and blind spot information #2 are both the same (001). In other words, blind spot information #1 and blind spot information #2 indicate blind spot information of divided blind spot areas obtained by dividing a blind spot area caused by one subject. The message may further include the identification number of the infrastructure sensor device 10, the coordinate information indicating the detectable range of the sensor unit 101, the time information indicating the time when the subject was detected, and the like, in addition to the subject information and the blind spot information.

The transmission unit 106 transmits the message generated by the message generation unit 105 to the communication device 11 of the vehicle via the communication network. Here, transmitting the message to the communication device 11 of the vehicle means, as shown in FIG. It also includes transmitting the message indirectly to the communication device 11 of the vehicle via another communication device. For example, the transmitter 106 may transmit a message to the base station via a wired or wireless communication network, and the base station may transmit the message to the communication device 11 of the vehicle via a wireless communication network. Alternatively, the transmission unit 106 may transmit the message to the communication device 11 via a server device (not shown). Further, when a message is directly transmitted from the transmission unit 106 to the communication device 11, it may be transmitted by broadcasting from the infrastructure sensor device 10 to the communication device 11 of a vehicle moving within a predetermined range. may be unicast to Note that the message generator 105 and the transmitter 106 correspond to the "transmitter" of the present invention.

3. Calculation of Planar Areas and Heights of Divided Blind Spot Areas Planar areas and heights of divided blind spot areas calculated by the blind spot calculator 104 of the present embodiment will be described with reference to FIGS. 4 to 7 .

FIG. 4a shows a simplified view of the road, objects on the road and the infrastructure sensor device 10 from above. As shown in FIG. 4a, a vehicle exists within the detection range of the sensor unit 101. FIG. Also, the oblique lines behind the vehicle indicate the blind area X caused by the vehicle, and FIG. 4a shows that there is a pedestrian in the blind area X. FIG. 4b shows an image detected by the sensor unit 101 of the infrastructure sensor device 10 of FIG. 4a. As described above, the position of the pedestrian is within the blind spot area X, but since the pedestrian is taller than the vehicle, the sensor unit 101 detects both the vehicle and the pedestrian behind the vehicle, as shown in FIG. can be detected.

FIG. 5 schematically shows the vehicle and the blind area viewed from the direction of the arrow shown in FIG. 4a. Figure 5a shows the three-dimensional blind area _X3D caused by the vehicle. Since the shaded area is the shadow of the vehicle when viewed from the infrastructure sensor device 10, it is a blind spot where the sensor unit 101 cannot detect the presence of an object. Although the shape of the three-dimensional blind spot area varies depending on the shape of the vehicle and the positional relationship between the vehicle and the infrastructure sensor device 10, the blind spot area having the shape of a triangular prism is assumed here for explanation.

The blind spot calculator 104 calculates a three-dimensional blind spot area as shown in FIG. 5a based on the subject information output from the subject detector 103 . Calculating the three-dimensional blind spot area may mean calculating the three-dimensional shape of the blind spot area. A region may be calculated.

FIG. 5b shows the planar area X, which is the projection of the three-dimensional blind area X _3D shown in FIG. 5a onto the ground, hatched. The blind spot calculator 104 calculates a planar area X of the blind spot area shown in FIG. 5b from the calculated three-dimensional blind spot area X _3D . For example, the blind spot calculator 104 calculates the planar area X by calculating the coordinates of the four vertices of the planar area X having a quadrangle.

Next, the blind spot calculation unit 104 obtains a position where a plane having a preset “predetermined” threshold height _ht and a three-dimensional blind spot area intersect, and projects this position onto the ground to obtain a figure. A plane region X shown in 5b is divided.

The term “predetermined” includes not only a case of being constant but also a case of being uniquely determined according to conditions.

FIG. 5c shows the blind spot area X as a three-dimensional split blind spot area a _3D (corresponding to a "first split blind spot area") having a height "greater than" a predetermined threshold height _ht set in advance; It shows a state divided into three-dimensional divided blind spot areas b _3D (corresponding to "second divided blind spot areas") having a height "below" the threshold height _ht . Assuming that the height of the vehicle is h _v , divided blind area a has a height of h _v to h _t , and divided blind area b has a height of h _t to 0. FIG. 5d shows planar areas a, b of the divided blind areas a _3D , b _3D shown in FIG. 5c, respectively.

"More than" and "less than" include both the case where the same value as the comparison object is included and the case where it is not included.

The blind spot calculation unit 104 transmits the height _hv , which is the maximum height of the divided blind spot region a _3D , and the planar region a to the message generation unit 105 as blind spot information indicating the divided blind spot region a. Similarly, the height _ht, which is the maximum height of the divided blind area b _3D , and the plane area b are transmitted to the message generator 105 as blind area information indicating the divided blind area b.

FIG. 6 also schematically shows the split blind areas a, b recognized by the vehicle receiving the above message. Figures 6a and 6b respectively show a side view and a top view of the divided blind area recognized by the vehicle. As described above, the message includes subject information indicating the vehicle and blind spot information indicating the divided blind spot areas a and b. It can be recognized that the indicated area is a blind spot area caused by the vehicle.

FIG. 7 further shows a state in which subject information indicating a pedestrian is displayed in the blind spot area shown in FIG. Similar to FIG. 6, FIGS. 7a and 7b respectively show the divided blind spot areas recognized by the vehicle and the subject viewed from the side and above.

When the sensor unit 101 detects a pedestrian positioned behind the vehicle, the vehicle receives subject information indicating the pedestrian in addition to subject information indicating the vehicle and blind spot information indicating the divided blind spot areas a and b. The position of the pedestrian overlaps with the position of the plane area b indicating that the sensor unit 101 cannot detect it, as shown in FIG. 7b. Therefore, when the height information is not included in the blind spot information, the vehicle receives inconsistent information that a pedestrian exists in the plane area b of the blind spot area that cannot be detected by the sensor unit.

On the other hand, in this embodiment, the blind spot information indicating the divided blind spot region b includes height _ht information in addition to the planar region b. Therefore, the blind spot information indicating the divided blind spot region b and the subject information indicating the pedestrian present at a position higher than the height _ht do not contradict each other. That is, since the pedestrian's upper body shown in FIG. 7a is outside the divided blind area b, even if the sensor unit 101 detects the pedestrian's upper body, it does not conflict with the blind area information indicated by the divided blind area b. Therefore, the vehicle can accurately recognize both the divided blind area b and the pedestrian.

The vehicle uses subject information indicating the vehicle and pedestrians and blind spot information indicating the divided blind spot areas a and b to calculate a travel plan for the vehicle. For example, when the vehicle travels near the divided blind areas a and b, it is possible to calculate a travel plan that decelerates the vehicle and takes a distance from the blind area. Furthermore, in this example, the vehicle can recognize that a pedestrian is present in the divided blind spot area b. This makes it possible to immediately detect a pedestrian running out of a blind spot.

The height included in the blind spot information is not limited to the maximum height of the divided blind spot area. For example, the height range of the divided blind area may be indicated by including the minimum height and the maximum height of the divided blind area in the blind area information. For example, in the example of FIG. 5, the maximum height _hv and the minimum height _ht may be transmitted as the height of the divided blind area a. Alternatively, the blind spot information may include the average value of the heights of the divided blind spot areas.

5 to 7 exemplify and explain the case where one threshold height is set in advance. However, multiple threshold heights may be preset. For example, threshold heights h _t1 , h _t2 may be set. In this case, the blind spot calculation unit 104 calculates the plane areas of the divided blind area with a height of _ht1 or more, the divided blind area with a height of _ht1 or less and _ht2 or more, and the divided blind area with a height of _ht2 or less. and height.

The predetermined threshold is set to 100 cm, for example, as a guideline for the height of preschool children. Alternatively, it may be set to 50 cm, which is lower than that. For example, if a child is in the blind spot created by the vehicle, the child may run out of the blind spot. Therefore, it is desirable that subject information when the subject is a child and blind spot information in which a child may be present are reliably used in calculating a vehicle travel plan. Therefore, by setting a value based on the child's height as the predetermined threshold value, the vehicle recognizes the blind spot information indicating the blind spot area and the subject information that detects the child existing in the blind spot area as contradictory information. can prevent it from being done.

4. Operation Next, the operation of the infrastructure sensor device 10 of the present embodiment will be described with reference to FIG.
First, the image acquisition unit 102 acquires an image from information detected by the sensor unit 101 (S101). When the subject detection unit 103 detects a subject in the image acquired in S101 (S102: Yes), the blind spot calculation unit 104 calculates a three-dimensional blind spot area caused by the subject, and divides the three-dimensional blind spot area into a plurality of blind spot areas. The planar area and height of each of the divided blind spot areas are calculated (S103). Then, the message generation unit 105 generates a message including subject information indicating the subject detected in S102 and blind spot information including the plane area and height calculated in S103 (S104). Then, the generated message is transmitted from the infrastructure sensor device 10 to vehicles traveling within a predetermined range (S105).

The timing at which the infrasensor device 10 executes the processing of FIG. 8 is arbitrary. For example, the infrastructure sensor device 10 may periodically execute the process of FIG. Alternatively, when an automatically driven vehicle passes within a predetermined range from the infrastructure sensor device 10, the process of FIG. 8 may be executed in response to a request from the automatically driven vehicle.

Next, the operation of the blind spot calculator 104 in S103 shown in FIG. 8 will be described in more detail with reference to FIG.
First, the blind spot calculation unit 104 calculates a three-dimensional blind spot area caused by the subject indicated by the subject information (S201). Next, the planar area of the blind area calculated in S201 is calculated (S202). Then, the position where the blind area and the plane of the threshold height intersect is determined (S203). The blind spot calculation unit 104 further divides the planar area of the blind area at the position where the intersecting position is projected onto the ground, and calculates the planar area of each of the divided blind areas (S204). The blind spot calculation unit 104 determines whether or not the flat areas and heights have been calculated for all blind spot areas (S205). If there is a blind spot area whose plane area and height have not been calculated (S205: No), the three-dimensional blind spot area is calculated again for the next blind spot area. Then, when the calculation of the planar area and height of the divided blind area is completed for all the blind area caused by the subject (S205: Yes), the blind area information including the planar area and height in addition to the subject information is sent to the message generation unit. 105 (S206).

Note that, in the present embodiment, an example in which the subject is a vehicle has been described. Various types of objects are assumed to exist on the road and its surroundings. There is a high possibility that an object such as a person exists. Therefore, the information on the blind spot area caused by the vehicle is useful information for the vehicle moving around the infrastructure sensor device 10 to travel safely. Therefore, it is desirable to apply the configuration of this embodiment to the calculation of the blind spot area caused by the vehicle. Therefore, when calculating the planar area and height of the divided blind area only for the blind area caused by the vehicle, before executing the series of processes shown in FIG. may be executed to determine whether or not to detect the blind spot area and to calculate the planar area and height of the divided blind spot area. However, this embodiment may be applied to a blind spot area caused by a subject other than a vehicle.

As described above, according to the present embodiment, the plane area and the height of each of a plurality of divided blind area areas obtained by dividing the blind area are transmitted to the vehicle as the blind area information. With the above configuration, even if the vehicle that communicates with the infrastructure sensor device receives both the blind spot information and the information on the object existing in the blind spot area indicated by the blind spot information, it recognizes that these pieces of information contradict each other. It is possible to calculate a travel plan of the vehicle using the blind spot information and the information of the object without doing so.

(Modification)
In the first embodiment described above, a configuration has been described in which the planar area and the height are calculated for all the divided blind area areas detected by the blind area calculation unit 104 and transmitted as blind area information. However, the blind spot calculation unit 104 may calculate the planar area and height of only some of the divided blind spot areas obtained by dividing the detected blind spot area, and transmit them as blind spot information.

For example, if the divided blind area has a certain height, there is a possibility that an object such as a pedestrian exists in this blind area. Therefore, the blind spot information indicating such a blind spot area is useful information for safe driving of the vehicle. On the other hand, if the height of the blind spot area is extremely low, there is little possibility that an object such as a pedestrian will be present in such a blind spot area to hinder the walking of the vehicle.

Therefore, the blind spot calculation unit 104 does not calculate a planar area for divided blind spot areas whose height is equal to or less than a predetermined value. Blind spot calculation section 104 calculates a planar area only when the height of the divided blind area is equal to or greater than a predetermined value, and outputs blind spot information including the calculated planar area and height to message generating section 105 . Therefore, the transmitting unit 106 transmits to the vehicle only the blind spot information of the divided blind spot areas whose height is equal to or greater than a predetermined value.

According to this modified example, blind spot areas that are low in height and less likely to affect the running of the vehicle can be excluded from the blind spot information to be transmitted to the vehicle. This makes it possible to reduce the amount of communication between the infrastructure sensor device 10 and the vehicle, in addition to reducing the processing in the blind spot calculation unit 104 .

(Embodiment 2)
In the first embodiment, a configuration has been described in which plane regions of divided blind spot regions obtained by dividing a three-dimensional blind spot region are calculated using a preset threshold height. In the present embodiment, a configuration will be described in which the blind spot calculator 104 calculates the height and plane area of the divided blind spot areas using a technique different from that of the first embodiment.

The height of the vehicle varies depending on the vehicle type. For example, light vehicles, ordinary vehicles, and trucks have different vehicle heights. 10a and 10b show states in which a light vehicle and a truck are traveling within the detection range of the sensor unit 101 of the infrastructure sensor device 10, respectively. When the flat area is calculated using the same threshold value for the light vehicle with a low vehicle height shown in FIG. 10a and the truck with a high vehicle height shown in FIG. 10b, the size of the calculated flat area is different. In particular, the planar area a of the truck is larger than that of the light vehicle.

Here, when the blind spot information including the planar area a and the height _hv of the divided blind spot area a of the truck is transmitted to the vehicle, the vehicle is configured such that the divided blind spot area a is a rectangular parallelepiped with the height _hv . Recognize it as an area. However, the actual height of the divided blind area a is not constant. Therefore, the wider the planar region a, the wider the region having the error between the actual height of the divided blind spot region a and the height of the divided blind spot region a recognized by the vehicle. descend. Therefore, the accuracy of the height of the divided blind area a of the truck shown in FIG. 10b is lower than the accuracy of the height of the divided blind area a of the light vehicle shown in FIG. 10a. Therefore, the blind spot calculation unit 104 of the present embodiment calculates the height of the threshold based on the height of the subject, and sets an appropriate threshold according to the height of the subject.

The blind spot calculator 104 calculates the height of the subject that causes the blind spot area. Then, a value that is half the height of the object is set as a predetermined threshold height. Alternatively, when the height of the subject is high like a truck, the threshold height may be set to 1/3 and 2/3 of the height of the subject, respectively. Then, the planar area of each divided blind area obtained by dividing the blind area into 2 or 3 is calculated. According to this example, by setting the threshold height according to the height of the subject, the error between the actual height of the divided blind spot area and the height indicated by the blind spot information can be reduced, and the accuracy of the blind spot information can be improved. can be enhanced.

As still another example, the blind spot calculator 104 may divide the blind spot area according to the size of the planar area of the blind spot area. As shown in FIG. 10, the planar area of a truck with a high vehicle height is larger than that of a light vehicle with a low vehicle height. Therefore, the blind spot calculation unit 104 calculates the planar area of each of the divided blind spot areas by, for example, dividing the planar area of the entire blind spot area into halves. Then, the height at which the vertical plane that divides the plane area into half and the blind area intersect may be calculated as the height of the divided blind area. In this case, unlike the first embodiment and the example described above, the height is calculated after obtaining the planar regions of the divided blind spot regions. According to this example, by dividing the planar area according to the size of the planar area and then calculating the height, it is possible to reduce the error between the actual height of the divided blind area and the height indicated by the blind area information. It is possible to improve the accuracy of blind spot information.

As yet another example, the blind spot calculator 104 may calculate the plane area and height using a table containing predetermined plane area and height values. 11A and 11B are diagrams for explaining blind spot areas caused by vehicles at different distances from the infrastructure sensor device 10. FIG. Although the vehicles shown in FIGS. 11a and 11b have the same vehicle height, the size of the planar area of the blind spot area differs depending on the distance from the infrastructure sensor device 10. FIG. As shown in FIG. 11, as the distance from the infrasensor device 10 increases, the planar area of the blind spot area increases, and the accuracy of the blind spot information decreases.

However, if the distance from the infrastructure sensor device 10 to the vehicle and the height of the vehicle are the same, the size of the planar area of the blind spot area caused by the vehicle is substantially constant. Therefore, the blind spot calculation unit 104 may store in a memory (not shown) in advance values of the plane area and the height for each distance from the infrastructure sensor device 10 to the vehicle and for each height of the vehicle. Then, based on the position and height of the vehicle, which is the subject detected by the subject detection unit 103, the values of the blind area and height stored in the table are extracted, and these values are divided into the planes of the divided blind area. May be set to area and height values. According to this example, the plane area and height values are stored in advance according to the position and height of the vehicle, which is the subject, so that the blind spot calculation unit 104 calculates the height or plane area by calculation. Since the height and plane area of the divided blind area can be obtained without calculation, it is possible to suppress the processing amount of the blind area calculation unit.

(Embodiment 3)
In Embodiments 1 and 2, the infra-sensor device 10 includes the object detection unit 103 and the blind spot calculation unit 104, and the configuration in which the infra-sensor device 10 calculates the planar area and height of the blind spot area has been described. However, the present invention may be implemented in devices other than the infrasensor device 10 .

FIG. 12 is a diagram illustrating the infrasensor system 2 according to this embodiment. The infrastructure sensor system 2 of this embodiment includes a server device 20 (corresponding to an “information processing device”), an infrastructure sensor device 21 , and a communication device 22 . FIG. 12 shows an example in which the server device 20 is a so-called edge server arranged in an area close to the infrastructure sensor device 21 . However, the server device 20 is a sensor server that collects and processes information from a plurality of infrastructure sensor devices 21, and is directly or indirectly connected to the infrastructure sensor devices 21 via a wired or wireless communication network. good too.

The configurations of the server device 20 and the infrastructure sensor device 21 that configure the infrastructure sensor system 2 will be described with reference to FIG. 13 .

The server device 20 includes a receiving unit 201 (corresponding to an “image acquiring unit”), a subject detecting unit 103 , a blind spot calculating unit 104 , a message generating unit 105 and a transmitting unit 106 . The receiving unit 201 receives information detected by the sensor unit 211 of the infrastructure sensor device 21, which will be described later. If the information received by the receiving unit 201 is not video information, the server device 20 may further include a configuration for converting the received information to generate video.

The subject detection unit 103, the blind spot calculation unit 104, and the message generation unit 105 of the server device 20 perform the same processing as the components described in the first and second embodiments. The transmitter 106 then transmits the message generated by the message generator 105 to the communication device 22 . As in the first embodiment, the transmission unit 106 may directly transmit the message to the communication device 22 or indirectly transmit the message to the communication device 22 via a base station or the like.

Like the infrastructure sensor device 10 of the first and second embodiments, the infrastructure sensor device 21 includes a sensor unit (corresponding to an “infrasensor”) 211 that detects objects present on the road or around the road. However, the infrastructure sensor device 21 of this embodiment transmits the information detected by the sensor unit 211 to the server device 20 via the communication unit 212 without processing it.

In the present embodiment, the server device 20 calculates the plane area and the height of each of the divided blind area obtained by dividing the three-dimensional blind area. In this embodiment, the infrastructure sensor device 21 needs to transmit the information detected by the sensor unit 211 to the server device 20, so communication traffic between the server device 20 and the infrastructure sensor device 21 is generated. However, since it is conceivable that the server device 20 can use a device with higher computational performance than the infrastructure sensor device 21, the object and the blind spot area are detected precisely, and the plane area and height are calculated with higher accuracy. is possible.

(Summary)
The features of the information processing apparatus according to each embodiment of the present invention have been described above.

Since the terms used in each embodiment are examples, they may be replaced with synonymous terms or terms including synonymous functions.

The block diagrams used in the description of the embodiments are obtained by classifying and arranging the configuration of the device for each function. Blocks representing respective functions are realized by any combination of hardware or software. Moreover, since the block diagram shows the function, it can also be understood as disclosure of the invention of the method and the invention of the program for realizing the method.

Regarding the functional blocks that can be understood as the processing, flow, and method described in each embodiment, the order is changed unless there is a restriction such that one step uses the result of another step that precedes it. good too.

The terms 1st, 2nd, and Nth (N is an integer) used in each embodiment and claims are used to distinguish between two or more configurations and methods of the same type. , does not limit the order or superiority.

In each embodiment, an infrastructure sensor device or a server device is assumed as an information processing device that implements the present invention. However, a device that implements the present invention is not limited to an infrastructure sensor device or a server device, and any information processing device capable of processing information detected by a sensor unit is sufficient. Information processing devices include electronic control units (ECUs), semiconductor circuit elements, PCs (personal computers), and the like.

It is assumed that the infrastructure sensor device of each embodiment is installed along a road and detects objects moving on and around the road. It also includes an infrastructure sensor device that is installed at a location and detects an object other than a vehicle.

In each embodiment, a configuration has been described in which a message including object information and blind spot information is transmitted to an in-vehicle communication device, but the information processing device may transmit to any communication device.

The present invention can be realized not only by dedicated hardware having the configuration and functions described in each embodiment, but also by a program for realizing the present invention recorded in a recording medium such as a memory or a hard disk, and a program capable of executing the program. It can also be realized as a combination with general-purpose hardware having a dedicated or general-purpose CPU and memory.

A program stored in a non-transitional material recording medium of dedicated or general-purpose hardware (for example, an external storage device (hard disk, USB memory, CD/BD, etc.), or an internal storage device (RAM, ROM, etc.)) It can also be provided to dedicated or general-purpose hardware via a recording medium or via a communication line from a server without a recording medium. This allows us to always provide the latest features through program upgrades.

Although the information processing device of the present invention has been described as an infrastructure sensor device or a server device, it can be applied to any device that processes video.

10, 20 information processing device, 102 video acquisition unit, 103 subject detection unit, 104 blind spot calculation unit, 105 message generation unit, 106 transmission unit, 201 reception unit

Claims (5)

an image acquisition unit (102) for acquiring an image detected by the infrastructure sensor;
a subject detection unit (103) for detecting a subject included in the image;
A predetermined threshold value is calculated based on the height of the subject, and a divided blind spot area obtained by dividing a three-dimensional blind area caused by the subject into a plurality of blind area areas having a height equal to or higher than the predetermined threshold. a blind spot calculation unit (104) for calculating plane regions and heights of one divided blind spot region and a second divided blind spot region having a height equal to or less than the predetermined threshold;
a transmission unit (105, 106) for transmitting blind spot information including the planar area and the height of the divided blind spot area;
An information processing device (10, 20). The transmission unit transmits the blind spot information when the height of the divided blind spot area is equal to or greater than a predetermined value.
The information processing apparatus according to claim 1. The information processing device is a server device (20),
the image acquisition unit acquires the image from the infrastructure sensor via a communication network;
The information processing apparatus according to claim 1. Acquire the image detected by the infrastructure sensor,
Detecting a subject included in the video,
calculating a predetermined threshold based on the height of the subject;
A first divided blind area having a height equal to or higher than the predetermined threshold and a divided blind area having a height equal to or lower than the predetermined threshold, which are divided blind areas obtained by dividing the three-dimensional blind area caused by the subject into a plurality of blind areas. Calculate the planar area and height of each of the second divided blind spot areas ,
transmitting blind spot information including the planar area and the height of the divided blind spot area;
Information processing methods. An information processing program executed by an information processing device,
Acquire the image detected by the infrastructure sensor,
Detecting a subject included in the video,
calculating a predetermined threshold based on the height of the subject;
A first divided blind area having a height equal to or higher than the predetermined threshold and a divided blind area having a height equal to or lower than the predetermined threshold, which are divided blind areas obtained by dividing the three-dimensional blind area caused by the subject into a plurality of blind areas. Calculate the planar area and height of each of the second divided blind spot areas ,
transmitting blind spot information including the planar area and the height of the divided blind spot area;
Information processing program.

Images