JP7497398B2 - Information processing device, information processing method, and program - Google Patents

Description

The disclosed embodiments relate to an information processing device, an information processing method, and a program.

Conventionally, technology is known that detects the lanes that separate the left and right sides of the vehicle's lane based on camera images captured by an onboard camera (see, for example, Patent Document 1).

JP 2015-090584 A

However, there is room for further improvement in the above-mentioned conventional technology in terms of preventing false lane detection.

For example, there are many road markings other than lanes around intersections. Among these road markings, pedestrian crossings in particular have multiple lines parallel to the lanes close to the lanes, which may lead to false detection of lanes.

One aspect of the embodiment has been made in consideration of the above, and aims to provide an information processing device, an information processing method, and a program that can prevent erroneous lane detection.

An information processing device according to one aspect of an embodiment extracts, from a plurality of line elements captured in a camera image of an in-vehicle camera, line element groups including a plurality of line elements that are parallel to each other and whose start and end positions are aligned in the depth direction on the camera image, and determines that the line element group is a crosswalk if the number of line elements at regular intervals in the line element group is equal to or greater than a predetermined number and there are a predetermined number or more sets of line elements whose start point spacing is within a predetermined range based on the standard dimensions of a crosswalk .

According to one aspect of the embodiment, erroneous detection of lanes can be prevented.

FIG. 1 is a diagram (part 1) illustrating a lane detection method according to an embodiment. FIG. 2 is a diagram (part 2) illustrating the lane detection method according to the embodiment. FIG. 3 is a block diagram illustrating an example of the configuration of the in-vehicle device according to the embodiment. FIG. 4 is a block diagram showing an example of the configuration of the lane detection processing unit. FIG. 5 is an explanatory diagram of the extraction conditions for line element groups. FIG. 6 is an explanatory diagram of the conditions for determining a pedestrian crossing. FIG. 7 is a diagram (part 1) showing a modified example of the exclusion process. FIG. 8 is a diagram (part 2) showing a modified example of the exclusion process. FIG. 9 is a diagram showing a modified example of the exclusion process (part 3). FIG. 10 is a flowchart showing a process performed by the in-vehicle device according to the embodiment.

Below, embodiments of the information processing device, information processing method, and program disclosed in the present application will be described in detail with reference to the attached drawings. Note that the present invention is not limited to the embodiments described below.

In the following, the information processing device according to the embodiment will be described as an in-vehicle device 10 (see FIG. 3) mounted on a vehicle. The in-vehicle device 10 is, for example, a drive recorder. In the following, the information processing method according to the embodiment will be described as a lane detection method executed by the in-vehicle device 10.

In addition, in the following, when we say "parallel," we do not mean a strictly parallel state, but rather broadly include cases where the two are nearly parallel.

Furthermore, in the following, when we say "predetermined," it may be replaced with "predetermined." For example, a "predetermined value" may be replaced with a "predetermined value."

In the following, when we say "lane," we are referring to the dividing lines themselves. When referring to the area in which vehicles can travel between the dividing lines, rather than the dividing lines themselves, we will use the term "traffic lane."

First, an overview of the lane detection method according to the embodiment will be described with reference to Figs. 1 and 2. Figs. 1 and 2 are explanatory diagrams (part 1) and (part 2) of the lane detection method according to the embodiment.

The vehicle-mounted device 10 has a camera 11 (see FIG. 3), which is an on-board camera. The vehicle-mounted device 10 executes lane detection processing based on camera images captured by the camera 11. In the lane detection processing, the vehicle-mounted device 10 detects lanes, which are dividing lines that separate the left and right sides of the vehicle's lane in which the vehicle is traveling.

The in-vehicle system including the in-vehicle device 10 can realize various functions such as a lane change determination function, a lane departure prevention support function, and a lane keeping support function based on the lane detection results by the in-vehicle device 10 and the distance between the lane position estimated from the detected lane and the vehicle position.

In the lane detection process, the in-vehicle device 10 binarizes the pixels of the camera image and extracts line elements. A line element is, for example, a line segment that appears as a boundary between a lane marking and a road surface. In other words, a line element is a straight edge on the camera image. The in-vehicle device 10 then selects a reference lane from among the lane candidates, which are the extracted group of line elements, and detects the lane through an identity determination between frames of the camera image.

By the way, line elements also appear as boundary lines between parts of the road surface and parts of the lane markings other than lanes. For example, Figure 1 shows line elements #1 to #14, which are the edge parts of a pedestrian crossing extracted as line elements near an intersection.

In lane detection processing, it is possible to exclude some of the causes of false detection from the processing by determining the reliability and continuity of the lanes estimated up to that point. However, since multiple line elements parallel to the lanes in real space are extracted in a position close to the lane of the vehicle's driving lane as shown in Figure 1, it is difficult to exclude crosswalks from processing by determining their reliability and continuity alone, which may lead to false detection of lanes.

In other words, in the example of Figure 1, if a crosswalk can be detected from the characteristics of the arrangement of line elements #1 to #14, and line elements #1 to #14 within the crosswalk area ER that is considered to be the crosswalk can be excluded from processing, it will be possible to prevent erroneous detection of lanes due to the crosswalk.

Therefore, in the lane detection method according to the embodiment, the controller 13 (see FIG. 3) of the in-vehicle device 10 extracts line element groups based on the depth length, start point position, end point position, parallelism, and start point interval of each line element. Then, if each line element of the line element group meets all the requirements of the number of line elements, line element position, and start point interval that correspond to the characteristics of a pedestrian crossing, the controller 13 determines that each line element of the line element group is a pedestrian crossing.

Figure 2 shows a schematic representation of the camera image of the vehicle-mounted camera shown in Figure 1 projected onto the world coordinate system. The X-axis direction is the depth direction, and the positive direction of the X-axis is the direction away from the vehicle.

As shown in FIG. 2, in the lane detection method according to the embodiment, the controller 13 extracts line element groups based on the depth length DL, start point position SP, end point position EP, parallelism AD, and start point interval SI of each extracted line element (step S1).

One end of the line element closest to the vehicle is the starting point, and the other end is the end point. The controller 13 searches for line elements in order of the closest X-axis position of the starting point, and sets the line element with the closest X-axis position of the starting point as the reference line element. In the example of Figure 2, line element #7 is the reference line element.

In step S1, the controller 13 extracts line elements whose depth length DL is within a predetermined range, whose starting point is within a range of starting point position SP based on the starting point of the reference line element, whose ending point is within a range of ending point position EP based on the ending point of the reference line element, and whose parallelism AD with respect to the reference line element is close to parallel, and groups them into a line element group. Line element groups whose starting point interval SI, which is the distance between the starting points in the Y-axis direction, is less than a predetermined value are treated as a single line element.

Note that the start point being within the range of the start point position SP indicates that the positions of the start points of each line element are aligned to a certain extent. Similarly, the end point being within the range of the end point position EP indicates that the positions of the end points of each line element are aligned to a certain extent.

Then, if each line element of the line element group satisfies all the requirements of the number of line elements, the position of the line elements, and the spacing between the line elements (corresponding to the start point spacing SI), the controller 13 determines that it is a crosswalk (step S2). In step S2, the controller 13 determines that each line element in the line element group is a crosswalk if the number of line elements is equal to or greater than a predetermined value, the line elements are located within a certain range CW in the center of the vehicle's travel lane for a predetermined value or more, and the start point spacing SI is close to the standard dimension of a crosswalk.

Figure 2 shows an example in which line elements #1 to #14 have been determined to be pedestrian crossings through steps S1 and S2. Specific numerical examples for each of the above-mentioned requirements will be described later with reference to Figures 5 and 6.

Then, the controller 13 excludes each line element determined to be a pedestrian crossing from the lane detection process. A specific example of this exclusion process will be described later using Figures 7 to 9, etc.

In this way, in the lane detection method according to the embodiment, the controller 13 extracts a line element group based on the depth length DL, start point position SP, end point position EP, parallelism AD, and start point interval SI of each line element. Then, the controller 13 determines that a line element is a pedestrian crossing when each line element in the line element group satisfies all of the requirements for the number of line elements, the position of the line element, and the interval between the line elements.

That is, the controller 13 detects a crosswalk based on the characteristics of the arrangement of line elements. In addition, the controller 13 excludes each line element detected as a crosswalk from the target of lane detection processing.

Therefore, the lane detection method according to the embodiment can prevent erroneous detection of lanes due to pedestrian crossings. Below, a more specific description will be given of an example configuration of an in-vehicle device 10 that applies the above-mentioned lane detection method.

Figure 3 is a block diagram showing an example of the configuration of the in-vehicle device 10 according to the embodiment. Also, Figure 4 is a block diagram showing an example of the configuration of the lane detection processing unit 13b. Note that Figures 3 and 4 show only the components necessary to explain the features of this embodiment, and descriptions of general components are omitted.

In other words, each component shown in Figures 3 and 4 is a functional concept, and does not necessarily have to be physically configured as shown. For example, the specific form of distribution and integration of each block is not limited to that shown, and all or part of it can be functionally or physically distributed and integrated in any unit depending on various loads, usage conditions, etc.

In addition, in the explanation using Figures 3 and 4, the explanation of components that have already been explained may be simplified or omitted.

As shown in FIG. 3, the in-vehicle device 10 according to the embodiment has a camera 11, a memory unit 12, and a controller 13.

The camera 11 is equipped with an imaging element such as a CCD (Charge Coupled Device) or a CMOS (Complementary Metal Oxide Semiconductor), and captures an image of a predetermined imaging range using this imaging element. The camera 11 is attached to various locations on the vehicle, such as the windshield or dashboard, so as to capture an image of a predetermined imaging range in front of the vehicle.

The storage unit 12 is realized by a storage device such as a RAM (Random Access Memory) or a flash memory, and in the example of FIG. 3, stores condition information 12a.

The condition information 12a is information that defines the conditions for extracting line element groups and the conditions for determining whether a crosswalk exists, which are requirements for detecting a crosswalk from the characteristics of the arrangement of line elements. Details of these conditions will be described later using Figures 5 and 6.

The controller 13 corresponds to a so-called processor or control unit. The controller 13 is realized by a CPU (Central Processing Unit), MPU (Micro Processing Unit), or the like, executing a program according to an embodiment (not shown) stored in the storage unit 12 using the RAM as a working area. The controller 13 can also be realized by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array).

The controller 13 has an acquisition unit 13a and a lane detection processing unit 13b, and realizes or executes the information processing functions and actions described below.

The acquisition unit 13a acquires the camera image captured by the camera 11 and outputs it to the lane detection processing unit 13b.

The lane detection processing unit 13b executes lane detection processing based on the camera image captured by the camera 11. The lane detection processing unit 13b also outputs the results of the lane detection processing to the external device 50.

The external device 50 is a device that executes applications corresponding to various functions such as the lane change determination function, lane departure prevention support function, and lane keeping support function described above. The external device 50 is, for example, an ECU (Electronic Control Unit). Note that the in-vehicle device 10 and the external device 50 may be integrated.

As shown in FIG. 4, the lane detection processing unit 13b has a line element extraction unit 13ba, a crosswalk detection unit 13bb, an exclusion unit 13bc, a lane extraction unit 13bd, a frame processing unit 13be, and an output unit 13bf.

The line element extraction unit 13ba binarizes the camera image and extracts line elements. The crosswalk detection unit 13bb executes the crosswalk detection process in steps S1 and S2 described above. Here, the crosswalk detection process will be explained with reference to Figures 5 and 6, giving specific values for various conditions.

Figure 5 is an explanatory diagram of the extraction conditions for line element groups. Also, Figure 6 is an explanatory diagram of the conditions for determining a pedestrian crossing. The crosswalk detection unit 13bb extracts line element groups based on the depth length DL, start point position SP, end point position EP, parallelism AD, and start point interval SI of each line element extracted by the line element extraction unit 13ba.

At this time, as shown in FIG. 5, the crosswalk detection unit 13bb extracts line elements whose depth length DL is within 1.0 m to 9.0 m, whose start point is at a start point position SP within ±1.0 m of the reference line element, whose end point is at an end point position EP within ±5.0 m of the reference line element, and whose parallelism AD is within ±10 deg of the reference line element, and groups them into a line element group. Only one line element group with a start point interval SI of less than 0.07 is counted.

In addition, the crosswalk detection unit 13bb determines that a crosswalk exists if each line element in a line element group satisfies all of the requirements for the number of line elements, the position of the line elements, and the spacing between the line elements.

In this case, as shown in FIG. 6, the crosswalk detection unit 13bb determines that each line element in a line element group is a crosswalk if, for example, there are seven or more line elements, one or more line elements are located within ±1.0 m of the center of the vehicle's travel lane, and there are four or more pairs of line elements with a starting point interval SI of 0.45 ±0.15 m or less.

Note that a number of line elements, for example seven or more, indicates that there are a sufficient number of line elements, and is an effective requirement for distinguishing between multiple lines and road markings other than pedestrian crossings (arrows, etc.). The same applies to the requirement for the position of the line elements. Furthermore, the requirement for the spacing of the line elements is based on the standard dimensions of the white line width and spacing of a pedestrian crossing being approximately 0.45 m, and is an effective requirement for distinguishing between road markings other than pedestrian crossings (letters, arrows, etc.).

Returning to the explanation of FIG. 4, the exclusion unit 13bc excludes each line element determined to be a crosswalk by the crosswalk detection unit 13bb from the processing target of the lane detection process. At this time, the exclusion unit 13bc excludes, for example, the line element itself determined to be a crosswalk from the processing target of the lane detection process.

Alternatively, the exclusion unit 13bc sets a rectangular crosswalk area ER that surrounds the line element group that is determined to be a crosswalk (see FIG. 1), and excludes the crosswalk area ER from the processing target.

Further modified examples of the exclusion process by the exclusion unit 13bc are shown in Figs. 7 to 9. Figs. 7 to 9 are diagrams (part 1) to (part 3) showing modified examples of the exclusion process. Note that, to make the explanation easier to understand, in Figs. 8 and 9, of the line elements #1 to #14 of the line element group determined to be a pedestrian crossing, the line elements #2 to #13 other than the outermost line elements #1 and #14 are shown by dashed lines.

As shown in FIG. 7, the exclusion unit 13bc may exclude frames in which a pedestrian crossing is detected ("crosswalk detection frames" in the figure) from the frames subject to lane detection processing, for example, between frames of a camera image.

Also, as shown in FIG. 8, the exclusion unit 13bc may process the rectangular pedestrian crossing area ER shown in FIG. 1 into a trapezoidal shape cut off by the outermost line elements in the line element group (line elements #1 and #14 in the example of FIG. 8), and exclude the trapezoidal area from the processing target.

As shown in FIG. 9, the exclusion unit 13bc may set the area surrounded by the line connecting the start points of the line element group determined to be a crosswalk, the line connecting the end points, and the outermost line element as the crosswalk area ER, and exclude the crosswalk area ER from the processing target. Note that in the explanation of the exclusion process, examples were given in which the crosswalk area ER was rectangular or trapezoidal, but the shape is not limited to these and may be any shape.

Returning to the explanation of FIG. 4, the lane extraction unit 13bd selects a reference lane based on the line elements extracted by the line element extraction unit 13ba and not excluded from processing by the exclusion unit 13bc. If the lane is a dashed line that indicates "overtaking" or "extraordination" is permitted, the lane extraction unit 13bd searches for such dashed lines and merges them into a single solid line to select the reference lane.

The frame processing unit 13be determines whether the lanes are the same between frames of the camera image. The output unit 13bf estimates the distance between the lane position of the detected lane and the vehicle position, the lateral movement speed, etc., and outputs them to the external device 50 as the results of the lane detection process.

Next, the processing procedure executed by the in-vehicle device 10 will be described with reference to FIG. 10. FIG. 10 is a flowchart showing the processing procedure executed by the in-vehicle device 10 according to the embodiment. Note that FIG. 10 shows the processing procedure for one frame of a camera image. This processing procedure is repeated while the lane detection process is being executed.

As shown in FIG. 10, the controller 13 of the in-vehicle device 10 acquires a camera image from the camera 11 (step S101).

Then, the controller 13 extracts line elements from the camera image (step S102). Next, the controller 13 extracts line element groups based on the depth length DL, start point position SP, end point position EP, parallelism AD, and start point interval SI of each line element (step S103).

Then, the controller 13 determines whether each line element in the line element group satisfies all of the requirements for the number of line elements, the position of the line elements, and the spacing between the line elements (step S104).

If all of the requirements are met (step S104, Yes), the controller 13 determines that the line element group is a crosswalk (step S105). The controller 13 then excludes each line element determined to be a crosswalk from the lane detection process (step S106), and proceeds to step S107.

On the other hand, if none of the requirements are met (step S104, No), the process proceeds directly to step S107. The controller 13 then executes lane detection processing in step S107, and ends processing of one frame of camera image.

As described above, the in-vehicle device 10 (corresponding to an example of an "information processing device") according to the embodiment includes the controller 13. The controller 13 executes lane detection processing to detect lanes based on line elements extracted from a camera image of the camera 11 (corresponding to an example of an "in-vehicle camera"). The controller 13 also extracts line element groups based on the depth length L, start point position SP, end point position EP, parallelism AD and start point interval SI of each line element, and if each line element of the line element group satisfies all of the requirements for the number of line elements, the position of the line elements and the interval between the line elements that correspond to the characteristics of a crosswalk, it determines that each line element of the line element group is a crosswalk.

Therefore, according to the in-vehicle device 10 of the embodiment, it is possible to exclude each line element determined to be a crosswalk from the target of the lane detection process, thereby preventing erroneous detection of lanes.

The controller 13 also extracts linear edges on the camera image as line elements, and extracts the line element whose starting point, which is the end on the front side in the depth direction on the camera image, is closest to the vehicle as the reference line element.

Therefore, according to the in-vehicle device 10 of the embodiment, each line element of the line element group can be extracted based on the reference line element.

The controller 13 also extracts line elements from the line element group whose starting point position SP is within a predetermined range based on the starting point of the reference line element.

Therefore, according to the in-vehicle device 10 of the embodiment, each line element whose starting point position is relatively aligned can be extracted as a line element of a line element group.

The controller 13 also extracts line elements from the line element group whose end point position EP is within a predetermined range based on the end point of the reference line element.

Therefore, according to the in-vehicle device 10 of the embodiment, each line element whose end point positions are relatively aligned can be extracted as a line element of a line element group.

The controller 13 also extracts line elements from the line element group whose depth length DL is within a predetermined range.

Therefore, according to the in-vehicle device 10 of the embodiment, each line element having a relatively uniform depth length DL can be extracted as a line element of a line element group.

The controller 13 also extracts line elements from the line element group whose angle difference with the reference line element indicating the parallelism AD is within a predetermined range.

Therefore, according to the in-vehicle device 10 of the embodiment, each line element that is nearly parallel to the reference line element can be extracted as a line element of the line element group.

In addition, the controller 13 extracts, for the line element group, multiple line elements whose start point interval SI, which is the horizontal interval between the start points, is less than a predetermined value as a single line element.

Therefore, the in-vehicle device 10 according to the embodiment can reduce the increase in processing costs caused by variations in line elements.

In addition, the controller 13 determines that the requirement regarding the number of line elements is met if the number of line elements in the line element group is equal to or greater than a predetermined value.

Therefore, the in-vehicle device 10 according to the embodiment can detect a crosswalk by distinguishing between multiple lines and road markings other than a crosswalk (arrows, etc.) and a crosswalk, based on the requirements regarding the number of line elements.

In addition, the controller 13 determines that the requirements regarding the position of the line element are met when the line element of the line element group is within a certain range of the center of the lane in which the vehicle equipped with the camera 11 is traveling by a predetermined value or more.

Therefore, the in-vehicle device 10 according to the embodiment can detect a crosswalk by distinguishing between multiple lines and road markings other than a crosswalk (arrows, etc.) and a crosswalk, based on requirements regarding the position of the line elements.

In addition, the controller 13 determines that the line element group satisfies the requirements regarding the spacing between the line elements if there are a predetermined number or more pairs of line elements whose start point spacing SI is within a predetermined range based on the standard dimensions of the crosswalk.

Therefore, the in-vehicle device 10 according to the embodiment can detect a crosswalk by distinguishing between a crosswalk and markings other than crosswalk marks and road markings (letters, etc.) based on the requirements regarding the spacing of line elements. In addition, the in-vehicle device 10 according to the embodiment can quickly obtain a crosswalk detection result by reusing the image processing results in the extraction of line elements, which has the secondary effect of reducing costs when the detection results are applied to other image recognition functions, etc.

The lane detection method according to the embodiment is an information processing method executed by the in-vehicle device 10, and includes executing a lane detection process to detect lanes based on line elements extracted from a camera image from the camera 11, extracting line element groups based on the depth length L, start point position SP, end point position EP, parallelism AD and start point interval SI of each line element, and determining that each line element of the line element group is a crosswalk if each line element of the line element group satisfies all of the requirements of the number of line elements, line element position and line element interval corresponding to the characteristics of a crosswalk.

Therefore, according to the lane detection method of the embodiment, it is possible to exclude each line element determined to be a pedestrian crossing from the target of the lane detection process, thereby preventing erroneous detection of lanes.

The program according to the embodiment also causes the computer to execute a lane detection process to detect lanes based on line elements extracted from the camera image of the camera 11, extract line element groups based on the depth length L, start point position SP, end point position EP, parallelism AD and start point interval SI of each line element, and determine that each line element in the line element group is a crosswalk if each line element in the line element group satisfies all of the requirements for the number of line elements, the position of the line element and the interval between the line elements that correspond to the characteristics of a crosswalk.

Therefore, according to the program of the embodiment, it is possible to exclude each line element determined to be a crosswalk from the target of the lane detection process, thereby preventing erroneous detection of lanes.

The program according to the embodiment can be recorded on a computer-readable recording medium such as a hard disk, a flexible disk (FD), a CD-ROM, a magneto-optical disk (MO), a digital versatile disk (DVD), or a universal serial bus (USB) memory, and can be executed by being read from the recording medium by a computer. Such a recording medium is also one aspect of the present disclosure.

The parameter values in the line element group extraction conditions and crosswalk determination conditions given in the above embodiment are merely examples and are not intended to be limiting.

Further advantages and modifications may readily occur to those skilled in the art. Therefore, the invention in its broader aspects is not limited to the specific details and representative embodiments shown and described above. Accordingly, various modifications may be made without departing from the spirit or scope of the general inventive concept as defined by the appended claims and equivalents thereof.

REFERENCE SIGNS LIST 10 In-vehicle device 11 Camera 12 Storage unit 12a Condition information 13 Controller 13a Acquisition unit 13b Lane detection processing unit 13ba Line element extraction unit 13bb Pedestrian crossing detection unit 13bc Exclusion unit 13bd Lane extraction unit 13be Frame processing unit 13bf Output unit 50 External device

Claims (10)

extracting a line element group including a plurality of line elements parallel to each other and having start and end positions aligned in a depth direction on the camera image from a plurality of line elements captured in a camera image of an in-vehicle camera;
If the number of line elements at regular intervals in the line element group is equal to or greater than a predetermined number, and the number of sets of line elements whose starting point intervals are within a predetermined range based on the standard dimensions of a pedestrian crossing is equal to or greater than a predetermined value , the line element group is determined to be a pedestrian crossing.
Information processing device. Extracting linear edges on the camera image as line elements;
extracting a line element whose starting point, which is one end on the front side in a depth direction on the camera image, is closest to the vehicle as a reference line element;
The information processing device according to claim 1 . extracting line elements whose starting point positions are within a predetermined range based on the starting point of the reference line element from the line element group;
The information processing device according to claim 2 . extracting line elements whose end point positions are within a predetermined range based on an end point of the reference line element from the line element group;
The information processing device according to claim 2 . extracting line elements whose depth length is within a predetermined range from the line element group;
The information processing device according to claim 1 . extracting line elements from the line element group, the angle difference between the line element and the reference line element indicating parallelism being within a predetermined range;
The information processing device according to claim 2 . extracting, from the line element group, a plurality of line elements whose start point interval, which is a horizontal interval between the start points, is less than a predetermined value as one line element;
The information processing device according to claim 2 . The line elements at regular intervals are line elements whose start point intervals are within a predetermined range based on the standard dimensions of the pedestrian crossing.
The information processing device according to claim 7. An information processing method executed by an information processing device,
extracting a line element group including a plurality of line elements parallel to each other and having start and end positions aligned in a depth direction on the camera image from a plurality of line elements captured in a camera image of an in-vehicle camera;
In the line element group, when the number of line elements at regular intervals is equal to or greater than a predetermined number and the number of sets of line elements whose start point intervals are within a predetermined range based on the standard dimensions of a pedestrian crossing is equal to or greater than a predetermined value , the line element group is determined to be a pedestrian crossing;
An information processing method comprising: extracting a line element group including a plurality of line elements parallel to each other and having start and end positions aligned in a depth direction on the camera image from a plurality of line elements captured in a camera image of an in-vehicle camera;
When the number of line elements at regular intervals in the line element group is equal to or greater than a predetermined number and the number of sets of line elements whose start point intervals are within a predetermined range based on the standard dimensions of a pedestrian crossing is equal to or greater than a predetermined value , the line element group is determined to be a pedestrian crossing;
A program that causes a computer to execute the following.

Images