JP2019194909A - Detection device, detection method and detection program - Google Patents

Abstract

To allow a driver to pay attention in advance.SOLUTION: A detection device of a first embodiment detects at least a sign in a region, a person and the region from an image including the region in which a moving body is movable, obtained by an acquisition unit. When the person exists in the region as a result of the detection, the device calculates a first distance from a detection position of the person to the person. The device sets a range according to the detection result of the sign in the region, the person and the region from the image, and the first distance.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a detection device, a detection method, a detection program, and an information processing system.

  A large percentage of pedestrians who suffer from traffic accidents are pedestrians who cross areas other than safety zones such as pedestrian crossings. When the driver travels in the vicinity of the safety zone, he / she empirically pays attention to the pedestrian. On the other hand, when the driver travels outside the safe zone, the driver is generally not consciously aware of the pedestrians. In this case, the driver must deal with a sudden crossing outside the safe zone. Is difficult. For this reason, conventionally, there has been proposed a technique for recording a case that is highly likely to cause an accident while the host vehicle is running, using an image taken by an in-vehicle camera.

JP 2011-134207 A JP 2005-157564 A

Tomoki Watanabe, Satoshi Ito and Kentaro Yokoi: “Co-occurrence Histograms of Oriented Gradients for Human Detection”, IPSJ Transactions on Computer Vision and Applications, Vol. 2, pp.39-47. (2010).

  The problem with conventional technology is that it is not possible to respond to sudden crossers outside of the safety zone, because it is possible to encounter accidents or to deal with crossers in pedestrian crossings. was there.

  The problem to be solved by the present invention is to provide a detection device, a detection method, a detection program, and an information processing system that can alert a driver in advance.

  The detection apparatus according to the first embodiment detects at least a label on a region, a person, and a region from an image including a region where the moving body is movable, which is acquired by the acquisition unit. When a person exists in the region, a first distance from the person detection position to the person is calculated. A range is set according to the indication on the region from the image, the detection result of the person and the region, and the first distance.

FIG. 1 is a block diagram illustrating an example of a hardware configuration of a detection apparatus that can be commonly applied to each embodiment. FIG. 2 is a diagram illustrating an example of a captured image captured by the camera according to each embodiment. FIG. 3 is a functional block diagram illustrating an example of functions of the detection device according to the first embodiment. FIG. 4 is a flowchart of an example showing the detection process according to the first embodiment. FIG. 5 is a flowchart illustrating an example of a human detection process according to the first embodiment. FIG. 6 is a diagram for explaining a distance calculation method using the installation posture of one camera according to the first embodiment. FIG. 7 is a diagram illustrating an example of the shape of the crossing person existence range according to the first embodiment on a two-dimensional map. FIG. 8 is a diagram for explaining an example of processing for changing the size of the crossing person presence range based on the distance relationship between the marking and the crossing person according to the first embodiment. FIG. 9 is a functional block diagram of an example for explaining the function of the detection device according to the modification of the first embodiment. FIG. 10 is a functional block diagram of an example for explaining the function of the detection device according to the second embodiment. FIG. 11 is a flowchart illustrating an example of detection processing according to the second embodiment. FIG. 12 is a diagram for explaining the positional relationship between the crossing person and the stopped vehicle in the captured image. FIG. 13 is a diagram for explaining a positional relationship in a captured image between a crossing person and a stopped vehicle. FIG. 14 is a diagram illustrating a configuration of an example of an information processing system according to the third embodiment. FIG. 15 is a functional block diagram illustrating an example of the function of the detection device according to the third embodiment. FIG. 16 is a functional block diagram of an example for explaining the functions of the server device according to the third embodiment. FIG. 17 is a flowchart illustrating an example of processing in the detection apparatus according to the third embodiment. FIG. 18 is a diagram illustrating an example of a crossing person presence range displayed on the detection device according to the third embodiment.

  Hereinafter, a detection device, a detection method, a detection program, and an information processing system according to the embodiments will be described.

  The detection device, the detection method, the detection program, and the information processing system according to the embodiment include, for example, a road that is a moving area of the host vehicle and a crosser that is a person who crosses the road from an image captured by an on-board camera. Detecting and setting a crossing person existence range where it is predicted that a crossing person will exist in the future according to the position where the crossing person is detected. By using the crossing area, it is possible to alert the driver to crossing persons who cross a safety zone such as a pedestrian crossing.

  FIG. 1 shows an example of a hardware configuration of a detection apparatus that can be commonly applied to each embodiment. In FIG. 1, a detection apparatus 10 includes a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a storage 104, an input / output I / F 105, a communication I / F. F106, the display control part 107, the position acquisition apparatus 108, and camera I / F109 are provided, and these each part is connected by the bus | bath 100 so that communication is possible.

  The CPU 101 controls the overall operation of the detection apparatus 10 by operating the RAM 103 as a work memory according to a program stored in advance in the ROM 102 or the storage 104. That is, a computer including the CPU 101 is mounted on the detection device 10. The storage 104 is a hard disk drive or a nonvolatile semiconductor memory (flash memory), and stores a program for operating the CPU 101 and various data.

  The input / output I / F 105 is, for example, a USB (Universal Serial Bus), and is an interface for transmitting / receiving data to / from an external device. Input devices such as a keyboard and a pointing device (such as a mouse) can be connected to the input / output I / F 105. Further, a drive device that reads a disk storage medium such as a CD (Compact Disk) or a DVD (Digital Versatile Disk) may be connected to the input / output I / F 105. The display control unit 107 converts the display control signal generated according to the program by the CPU 101 into a display signal that can be displayed on the display 120 using an LCD (Liquid Crystal Display) or the like as a display device, and outputs the display signal.

  The communication I / F 106 controls communication with a network such as a LAN (Local Area Network) or the Internet.

  The position acquisition device 108 acquires position information indicating the current position. The position information is expressed using, for example, latitude and longitude. The location information may further include an altitude. The position acquisition device 108 acquires the current position using, for example, a GNSS (Global Navigation Satellite System). However, the present invention is not limited to this, and the position acquisition device 108 may acquire the current position using a wireless LAN position information system or a self-position estimation system using an in-vehicle sensor.

  The camera 110 captures an image according to the control of the camera I / F 109 and outputs an image signal. The camera I / F 109 controls the camera 110 in accordance with an instruction from the CPU 101, captures an image signal output from the camera 110, and outputs it as a captured image.

  The camera 110 is mounted, for example, so that a moving area in which the host vehicle can travel in the traveling direction of the host vehicle is included in the captured image. FIG. 2 shows an example of a captured image captured by the camera 110 according to each embodiment. In FIG. 2, a captured image 200 includes an image of a moving area 201 that is a road on which the host vehicle is traveling, for example. In the example of FIG. 2, the captured image 200 includes a sign 202 that is a sign on the moving area 201, a person 203 that crosses the moving area 201, and an image of another vehicle 204 on the moving area 201. . In the example of FIG. 2, the sign 202 indicates a pedestrian crossing.

  The camera 110 is not limited to imaging light in the visible light region, but may be an infrared camera that can image light in the infrared region or an ultraviolet camera that can image light in the ultraviolet region. In FIG. 1, one camera 110 is shown to be connected to the detection apparatus 10, but this is not limited to this example. For example, a plurality of cameras 110 may be connected to the detection device 10 and each camera 110 may be installed in different directions, for example, in front, rear, and side of the host vehicle. In the following, it is assumed that a single camera 110 installed so as to image the traveling direction of the host vehicle is connected to the detection device 10.

(First embodiment)
A first embodiment will be described. FIG. 3 is a functional block diagram illustrating an example of functions of the detection device according to the first embodiment. In FIG. 3, the detection device 10 a includes an image acquisition unit 11, a detection unit 12, a calculation unit 13, a setting unit 14, and a display unit 15. The image acquisition unit 11, the detection unit 12, the calculation unit 13, the setting unit 14, and the display unit 15 are realized by a detection program that operates on the CPU 101. Not limited to this, some or all of the image acquisition unit 11, the detection unit 12, the calculation unit 13, the setting unit 14, and the display unit 15 may be configured by hardware circuits that operate in cooperation with each other.

  The image acquisition unit 11 acquires a captured image captured by the camera 110. The detection unit 12 detects at least an image of the person 203, an image of the moving area 201, and an image of the sign 202 on the moving area 201 from the captured image acquired by the image acquisition unit 11. In the following, “acquiring an image of a person 203” is described as appropriate, for example, “acquiring a person 203”.

  When the detection unit 12 detects the person 203 and the moving area 201 and determines that the detected person 203 exists on the moving area 201, the calculation unit 13 determines that the detected person 203 is a crossing person. Assuming that there is a crossing distance, which is a distance from the own vehicle to the crossing, based on the captured image. The setting unit 14 sets a crossing person existence range in which it is estimated that a crossing person exists by using the detection result by the detection unit 12 and the crossing person distance calculated by the calculation unit 13.

  The display unit 15 causes the display 120 to display the crossing person presence range set by the setting unit 14. For example, the display unit 15 can display the captured image 200 on the display 120 and display the captured image 200 by overlapping the crossing person presence range. Not limited to this, the display unit 15 may display a map including the current position based on the current position acquired by the position acquisition device 108 and display the crossing person presence range on the map.

  FIG. 4 is a flowchart of an example showing the detection process according to the first embodiment. In step S <b> 100, the image acquisition unit 11 acquires a captured image captured by the camera 110. When a captured image is acquired by the image acquisition unit 11 in step S100, the process proceeds to step S101 and step S102.

  In step S101, the detection unit 12 detects the person 203 from the captured image acquired in step S100. In step S102, the detection unit 12 acquires the movement area 201 from the captured image acquired in step S100. When the person 203 and the moving area 201 are detected in step S101 and step S102, the process proceeds to step S103.

  In FIG. 4, the processing of step S101 and step S102 is shown to be executed in parallel. However, this is not limited to this example, and step S101 and step S102 may be executed sequentially. In this case, the order of step S101 and step S102 may be arbitrary.

  In step S103, the calculation unit 13 determines whether the person 203 exists on the moving area 201 when both the person 203 and the moving area 201 are detected in steps S101 and S102. When the calculation unit 13 determines that the person 203 exists on the moving area 201, the calculation unit 13 determines that the person 203 is a crossing person (Yes in Step S103) and shifts the process to Step S104.

  On the other hand, in step S103, the calculation unit 13 detects that at least one of the person 203 and the moving area 201 is not detected in steps S101 and S102, or the person 203 and the moving area 201 are detected in steps S101 and S102. When the detected person 203 exists outside the moving area 201 (No at Step S103), a series of processes according to the flowchart of FIG.

  In step S104, the calculation unit 13 calculates the distance from the own vehicle (camera 110) to the crossing person (person 203) based on the captured image 200 and the installation angle of the camera 110 with respect to the horizontal. In the next step S105, the setting unit 14 sets a crossing person existence range, which is a range in which it is estimated that a crossing person exists, using the distance to the crossing person calculated in step S104. The set crossing person presence range can be displayed on the display 120 together with the captured image 200, for example.

  Next, each process in the flowchart of FIG. 4 will be described in more detail. First, a process of detecting an image of the person 203 included in the captured image 200 by the detection unit 12 in step S101 will be described. For the detection of the person 203, an evaluation value representing humanity is calculated, and when the evaluation value is equal to or greater than a set threshold value, it is determined that the person is a person.

  FIG. 5 is a flowchart illustrating an example of the detection process of the person 203 by the detection unit 12 according to the first embodiment. In step S1010, the detection unit 12 sets a detection range for performing human detection on the captured image 200 acquired in step S100. The detection range may be, for example, the entire captured image 200 or a part of the captured image 200. For example, a region excluding the upper end portion and the lower end portion of the captured image 200 can be set as the detection range. Alternatively, a mask image may be prepared in advance, and the setting area may be designated by the mask image. The detection unit 12 sets a detection window at a predetermined position within the detection range. The size and shape of the detection window are not particularly limited. In the following description, it is assumed that the detection window is rectangular.

  In the next step S1011, the detection unit 12 determines whether or not the processing within the detection range has ended. If it is determined that the process has been completed (step S1011: Yes), a series of processes according to the flowchart of FIG. If it is determined that the process has not been completed (No in step S1011), the process proceeds to step S1012.

  In step S1012, the detection unit 12 calculates a feature amount in the detection window region set in the captured image 200. The detection unit 12 can use, for example, an HOG (Histograms of Oriented Gradients) feature quantity as the feature quantity. In step S1013, the detection unit 12 calculates an evaluation value representing humanity using a discriminator based on the feature amount calculated in step S1011. The detection unit 12 can use, for example, an SVM (Support Vector Machine) as an identifier.

  In the next step S1014, the detection unit 12 determines whether or not the evaluation value calculated in step S1012 is greater than or equal to a preset threshold value. If the detection unit 12 determines that the evaluation value is less than the threshold (No in step S1014), the detection unit 12 shifts the process to step S1016.

  On the other hand, if the detection unit 12 determines in step S1014 that the evaluation value is equal to or greater than the threshold value (step S1014, Yes), the detection unit 12 shifts the process to step S1015 and determines that the person 203 is included in the detection window. Specifically, for example, using the technique disclosed in Non-Patent Document 1, the position of the detection target is detected. After the process of step S1015, the process proceeds to step S1016.

  In step S1016, the detection unit 12 moves the detection window and returns the process to step S1011.

  Next, processing for detecting an image of the moving area 201 included in the captured image 200 by the detection unit 12 in step S102 in the flowchart of FIG. 4 will be described. The detection unit 12 detects the movement area 201 by, for example, obtaining an installation posture of the camera 110 in advance and setting an area in which the road surface is reflected when the width of the road surface (movement area 201) is assumed to be constant. It can be calculated based on the posture. The installation posture includes, for example, the depression angle of the camera 110, the angle of the imaging direction of the camera 110 with respect to the traveling direction of the host vehicle, and the installation height of the camera 110. Not limited to this, the detection unit 12 presets an area where the moving area 201 is included in the captured image 200 from the installation position of the camera 110, detects a straight line within the set area by Hough transform, and the vanishing point. An area surrounded by two straight lines passing through the vicinity may be detected as the moving area 201.

  Next, a description will be given of the process of determining whether or not the person 203 exists in the moving area 201 by the calculation unit 13 in step S103 in the flowchart of FIG. The calculating unit 13 determines whether or not the person 203 exists in the moving area 201 using the person 203 and the moving area 201 detected by the detecting unit 12 in step S101 and step S102. If it is determined that it exists, the person 203 is regarded as a crossing person.

  For example, the calculation unit 13 determines whether the person 203 moves to the moving area 201 based on whether or not the lower end of the detection window determined by the detecting unit 12 in step S1015 of the flowchart of FIG. It can be determined whether or not it exists inside. In this example, when the lower end of the detection window is included in the moving area 201, it can be determined that the person 203 exists in the moving area 201.

  Further, for example, in the detection window determined that the person 203 is included in step S1015, the foot area of the person 203 is further detected. Based on whether or not the detected foot area is included in the movement area 201, the person 203 It can be determined whether or not it exists in the movement area 201. In this example, when the foot area is included in the moving area 201, it can be determined that the person 203 exists in the moving area 201.

  The method described in the human detection can be used for detecting the foot region. For example, the detection unit 12 obtains a feature amount from the captured image 200, and calculates an evaluation value that represents the likelihood of the foot region using a discriminator based on the obtained feature amount. A foot region is detected by determining the threshold value of this evaluation value.

  Next, a process of calculating the distance between the person 203 and the host vehicle (camera 110) by the calculation unit 13 in step S104 in the flowchart of FIG. 4 will be described. For example, when the host vehicle has two or more cameras 110, the distance between the camera 110 and the crossing person (person 203) can be easily obtained by the stereo method.

  When the host vehicle is equipped with one camera 110, the camera 110 is used based on the foot position information indicating the position of the crossing person's foot area on the captured image 200 by using the installation posture of the camera 110. And the distance between crossers.

  A distance calculation method using the installation posture of one camera 110 according to the first embodiment will be described with reference to FIG. In FIG. 6, it is assumed that the camera 110 with a focal length f is installed at a height h from the road surface 500 with the imaging direction 503 as the depression angle θ. A straight line 502 intersecting at right angles to the imaging direction 503 at the position of the focal distance f indicates a virtual captured image 200 by the camera 110. The value p indicates the position of the foot region of the person 501 with reference to the center of the captured image 200, that is, the intersection of the imaging direction 503 and the straight line 502.

  In this case, the distance d from the position where the perpendicular from the camera 110 intersects the road surface 500 to the foot region of the person 501 can be calculated by the following equation (1).

  Next, the crossing person existence range setting process by the setting unit 14 in step S105 according to the flowchart of FIG. 4 will be described. In step S105, the setting unit 14 estimates that a crossing person exists from the detection result of the detection unit 12 and the distance d between the camera 110 and the crossing person (person 203) calculated by the calculation unit 13. Set the range. That is, the setting unit 14 sets a crossing person existence range which is a range in which it is estimated that the same or different crossing persons exist after that time, not the time when the crossing person is observed.

  In this way, by setting and storing the range in which the crossing can exist in the future, after the time when the crossing was detected, the driver should be alerted in advance before actually detecting the crossing. Can be aroused.

  First, the setting unit 14 sets a range of a preset size around a position where a crossing person is detected as a crossing presence range. The position where the crossing person is detected is, for example, the position of the distance d calculated in step S104 described above from the position of the camera 110 toward the imaging direction on the horizontal plane of the camera 110.

  FIG. 7 shows an example of the shape of the crossing person presence range according to the first embodiment on a two-dimensional map. FIG. 7A shows an example of a crossing person presence range 211a set in a circular shape with a predetermined radius with the crossing person detection position 210 as the center. FIG. 7B shows an example of a crossing person existence range 211b set in a rectangular shape (square in this example) having a predetermined side length with the crossing person detection position 210 as the center. FIG. 7C shows an example of a crossing person presence range 211c set in accordance with the shape of the road 212 with the crossing person detection position 210 as the center.

  As for the size of the crossing person existence range, for example, in the case of the circular crossing person existence range 211a shown in FIG. 7A, it is conceivable that the radius is about several times the width of the road. For example, if the width of the road is 10 m, it is conceivable that the radius of the crossing person existence range 211a is several tens of meters (for example, about 20 m to 30 m).

  In the first embodiment, the size of the crossing person presence range is further changed according to the detection result of the detection unit 12. As an example, the size of the crossing person presence range can be changed based on the distance relationship between the sign 202 (pedestrian crossing) detected from the captured image 200 and the crossing person.

  An example of processing for changing the size of the crossing person presence range based on the distance relationship between the marking 202 and the crossing person according to the first embodiment will be described with reference to the flowchart of FIG. In step S <b> 1051, the detection unit 12 detects, for example, a pedestrian crossing area as one of the signs 202 from the movement area 201.

  The detection unit 12 can use, for example, the method described in the human detection as a method for detecting the pedestrian crossing area. For example, the detection unit 12 obtains a feature amount from the captured image, and calculates an evaluation value representing the likelihood of a pedestrian crossing using a discriminator based on the obtained feature amount. A pedestrian crossing area is detected by determining the threshold value of this evaluation value. The feature amount is not limited to the above-described HOG feature amount, and a Gabor feature can also be used as a feature amount that can effectively use the brightness difference.

  In step S1052, the detection unit 12 determines whether the detection of the pedestrian crossing area has been successful. If it is determined that the detection of the pedestrian crossing area is successful (step S1052, Yes), the process proceeds to step S1053. On the other hand, when it is determined that the detection has failed (No at Step S1052), a series of processes according to the flowchart of FIG.

  In step S1053, the calculation unit 13 calculates the distance from the camera 110 to the pedestrian crossing area detected in step S1052. Since the sign 202 exists on the road surface, that is, on the moving area 201, this distance can be calculated using the above-described equation (1). In the next step S1054, the calculation unit 13 determines whether the sign 202 and the crossing person (person 203) are based on the distance calculated in step S104 in the flowchart of FIG. 4 and the distance to the pedestrian crossing area calculated in step S1053. A distance D between them is calculated.

  In the next step S1055, the setting unit 14 determines whether or not the distance D calculated in step S1054 is less than a threshold value. When the setting unit 14 determines that the distance D is greater than or equal to the threshold (No in step S1055), the setting unit 14 ends the series of processes according to the flowchart of FIG. On the other hand, if the setting unit 14 determines in step S1055 that the distance D is less than the threshold value (step S1055, Yes), the setting unit 14 shifts the process to step S1056.

  In step S1056, the setting unit 14 adjusts the size of the crossing person presence range according to the value of the distance D. In the first embodiment, the setting unit 14 reduces the crossing person presence range at a constant rate.

  For example, the setting unit 14 adjusts the distance D so that the size of the crossing person presence range is inversely proportional. In this case, for the circular crossing person existence range 211a shown in FIG. 7A, it is conceivable to adjust the distance D and the radius of the crossing person existence range 211a to be inversely proportional. Not limited to this, the setting unit 14 may make the size of the crossing person presence range inversely proportional to the square of the distance D. Furthermore, the setting unit 14 may expand the crossing person presence range according to the distance D when the distance D is equal to or greater than the threshold.

  As described above, according to the first embodiment, when a crossing person crossing the pedestrian crossing is detected on the road on which the host vehicle is traveling, a predetermined size is detected with respect to the detected position of the crossing person. The crossing presence range is set. By using the crossing person presence range set in this way, a range where the crossing person can exist can be presented to the driver, and the driver can be alerted.

(Modification of the first embodiment)
A modification of the first embodiment will be described. FIG. 9 is a functional block diagram of an example for explaining the function of the detection device according to the modification of the first embodiment. Note that, in FIG. 9, the same reference numerals are given to portions common to FIG. 3 described above, and detailed description thereof is omitted.

  In FIG. 9, the detection apparatus 10 a ′ has a position information acquisition unit 16 added to the detection apparatus 10 a illustrated in FIG. 3. The position information acquisition unit 16 acquires the current position using the position acquisition device 108. The position information acquisition unit 16 passes the acquired position information to the setting unit 14. Further, the position information acquisition unit 16 acquires the position where the captured image is acquired in response to the notification of the captured image acquisition from the image acquisition unit 11, and passes the position information indicating the position to the setting unit 14. For example, the setting unit 14 adds the position information passed from the position information acquisition unit 16 to the set crossing person existence range, and thus the crossing person existence range on the two-dimensional map as shown in FIG. The display can be performed on the display unit 15.

  Further, the set crossing person presence range information can be stored in the storage 104, for example, in association with the position information. In this case, when the host vehicle travels at a position close to the stored position information, a crossing person is not actually detected at that time, but there is a possibility that a crossing person exists. It can be presented to the driver, which can alert the driver.

(Second Embodiment)
Next, a second embodiment will be described. FIG. 10 is a functional block diagram of an example for explaining the function of the detection device according to the second embodiment. In FIG. 10, the same reference numerals are given to the portions common to FIG. 3 described above, and detailed description thereof is omitted.

  In FIG. 10, the detection apparatus 10b has an age estimation unit 21 and a stopped vehicle recognition unit 22 added to the detection apparatus 10a shown in FIG. The age estimating unit 21 recognizes the attribute information based on the image of the crossing person in the captured image 200 for the crossing person calculated by the calculating unit 13 in step S103 in FIG. 4, and determines the age of the crossing person from the recognized attribute information. Is estimated. In the detection device 10b, the detection unit 12 further detects the other vehicle 204 existing in the moving area 201 from the captured image 200 acquired by the image acquisition unit 11. The stop vehicle recognition unit 22 recognizes whether the other vehicle 204 detected by the detection unit 12 is stopped.

  The setting unit 14 changes the size of the set crossing person existence range according to the age of the crossing person estimated by the age estimation unit 21. Further, the setting unit 14 obtains the positional relationship between the other vehicle 204 recognized by the stopped vehicle recognition unit 22 and the crossing person obtained by the calculating unit 13, and the size of the crossing person existence range is increased according to the obtained positional relation. Change the size.

  FIG. 11 is a flowchart illustrating an example of detection processing according to the second embodiment. Note that the flowchart of FIG. 11 is executed after the crossing person presence range is set by the process of step S105 in the flowchart of FIG. 4 described above.

  In step S <b> 201, the age estimation unit 21 recognizes attribute information from the crosser's image for the crosser obtained by the calculation unit 13, and estimates the age of the crosser based on the recognized attribute information. In the next step S202, the stopped vehicle recognition unit 22 recognizes whether or not the other vehicle 204 detected by the detection unit 12 is a stopped vehicle. Note that the order of the processes in step S201 and step S202 may be changed, or the processes may be executed in parallel.

  In the next step S203, the setting unit 14 determines whether or not the age estimated by the age estimating unit 21 in step S201 is an age outside a predetermined age range. When the setting unit 14 determines that the estimated age is within the predetermined age range (step S203, No), the process proceeds to step S205. On the other hand, when the setting unit 14 determines that the estimated age is out of the predetermined age range (step S203, Yes), the process proceeds to step S204.

  In step S204, the setting unit 14 expands the crossing person presence range set in step S105 of the flowchart of FIG. Then, the process proceeds to step S205.

  The determination process in step S203 determines whether the crossing person is an elderly person or a younger person, and the predetermined age range is set to, for example, 10 years old or more and less than 60 years old. In this case, it is determined that the estimated age is less than 10 years old or 60 years old or more and is outside the predetermined age range, the process proceeds to step S204, and the crossing person presence range is expanded. Elderly and younger people move slower than other age groups (generally adults) (eg older and younger people), or movement is uncertain (eg younger people) It is thought that. Therefore, when it is estimated that the crossing person is an elderly person or a low-age person, the crossing person existence range is expanded in step S204, and the effect of prompting the driver's attention is improved.

  In step S205, the stopped vehicle recognition unit 22 determines whether the other vehicle 204 detected by the detection unit 12 is stopped according to the recognition result of step S202. When the stopped vehicle recognition unit 22 determines that the other vehicle 204 is not stopped, that is, is moving (No in step S205), the series of processes according to the flowchart of FIG. 11 is terminated. On the other hand, if the stopped vehicle recognition unit 22 determines that the detected other vehicle 204 is stopped (step S205, Yes), the process proceeds to step S206.

  In step S <b> 206, the setting unit 14 adjusts the size of the crossing person presence range based on the positional relationship between the other vehicle 204 recognized as being stopped and the crossing person on the captured image 200. Then, a series of processes according to the flowchart of FIG. 11 is terminated.

  Next, the age estimation process by the age estimation unit 21 in step S201 will be described. The age estimation unit 21 recognizes the crossing person's attribute information detected by the detection unit 12 based on the crossing person image. As the attribute information of the crossing person, for example, the gender, skin color, height, clothes, belongings, physical characteristics, etc. of the crossing person can be considered. The age estimation unit 21 estimates the age of the crosser based on the attribute information recognized for the crosser. For example, the age estimation unit 21 can estimate the age based on the height among the attribute information of the crossing person. In this case, it is conceivable that the crosser's height is equal to or less than a predetermined value, and that the crosser is an elderly person or an elderly person.

  The height of the crossing person can be recognized by applying the method in which the calculation unit 13 calculates the distance from the camera 110 to the crossing person in step S104 of FIG. For example, the crossing person's foot region and head region are recognized from the crossing person's image, and the recognized foot region and head region in the captured image 200 and the distance to the crossing person's foot region are determined. Recognize your height. As an example, a height threshold is set, and the height (number of pixels) of the threshold height in the captured image 200 is obtained in advance according to each distance. The height of the crosser can be calculated based on the height and distance from the foot region to the head region of the crosser.

  However, the present invention is not limited to this, and it is also possible to estimate the age of the crosser based on the moving speed of the crosser. In this case, it is conceivable that the crossing person is tracked in time series to obtain the moving speed, and when the obtained moving speed is equal to or less than a set threshold, it is determined that the crossing person is an elderly person or a low-age person.

  Next, the stop vehicle recognition process by the stop vehicle recognition unit 22 in step S202 will be described. The stopped vehicle recognition unit 22 tracks the other vehicle 204 detected by the detection unit 12 in time series, and determines whether the other vehicle 204 is moving or stopped based on the amount of movement. Similar to the process of step S104 in the flowchart of FIG. 4, the distance from the camera 110 to the other vehicle 204 can be calculated based on the lower end region of the other vehicle 204. Accordingly, the other vehicle 204 is tracked in time series, and the distance to the other vehicle 204 is sequentially obtained in time series, whereby the relative speed of the other vehicle 204 with respect to the own vehicle can be calculated. May be the amount of movement.

  Next, the process according to the positional relationship on the captured image 200 of the other vehicle 204 and a crossing person of step S206 is demonstrated. When the stop vehicle recognition unit 22 recognizes that the other vehicle 204 is stopped, the setting unit 14 determines the size of the crossing person existence range according to the distance between the other vehicle 204 and the crossing person. adjust. Here, the distance between the stopped other vehicle 204 and the crossing person may be a distance on the captured image 200, or may be an actual distance calculated as described above for each of the other vehicle 204 and the crossing person.

  For example, the setting unit 14 uses the crossing person's foot position on the captured image 200 and the lower end position of the stopped vehicle region to determine the positional relationship between the crossing person and the stopped vehicle in the captured image 200.

  The positional relationship in the captured image 200 of a crossing person and a stop vehicle is demonstrated using FIG. 12 and FIG. As illustrated in FIG. 12, the horizontal direction in the captured image 200 between the foot position 220 of the crossing person (person 203) detected in the captured image 200 and the lower end position 221 of the stopped vehicle (other vehicle 204). The difference between the coordinate values is defined as the width Δw, and the difference between the coordinate values in the vertical direction is defined as the height Δh. The setting unit 14 sets threshold values for the width Δw and the height Δh, respectively.

  FIG. 13A shows an example in which the width Δw is greater than or equal to the threshold value. FIG. 13B shows an example in which the height Δh is greater than or equal to the threshold value. The setting unit 14 determines that the distance between the other vehicle 204 and the crossing person in the real space is close when the width Δw and the height Δh are each equal to or less than the threshold value. In this case, the setting unit 14 expands the crossing person presence range, for example. For example, the setting unit 14 determines that the distance between the other vehicle 204 and the crossing person in the real space is long when the width Δw and the height Δh are each equal to or greater than the threshold value. In this case, the setting unit 14 reduces the crossing person existence range, for example. In other cases, the setting unit 14 does not adjust the crossing person presence range in step S206.

  In addition, the determination method of the positional relationship between a crossing person and a stop vehicle is not limited to the above-mentioned example. For example, when at least one of the height Δh and the width Δw is equal to or greater than a threshold value, it may be determined that the distance between the other vehicle 204 and the crossing person is long.

  As described above, in the second embodiment, the crossing person existence range set by the processing of the first embodiment is further adjusted based on the crossing person attribute information and the distance between the crossing person and the stopped vehicle. I am doing so. Therefore, the driver can be alerted more accurately.

(Third embodiment)
Next, a third embodiment will be described. FIG. 14 illustrates an exemplary configuration of an information processing system according to the third embodiment. 14, the information processing system includes a server device 60 including a database (DB) 61, and one or more detection devices 10c, 10c,... Are connected to the network 51 by wireless communication, and the server device 60 is connected via the network 51. Communicate with. The AP 50 is a relay device for connecting the detection devices 10c, 10c,... To the network 51 by wireless communication. The wireless communication performed by the detection device 10c is not particularly limited as long as data communication with a predetermined communication partner is possible, but a wireless LAN (Local Area Network) can be used.

  In the third embodiment, in such a configuration, each detection device 10c associates the set crossing person existence range information with the position information indicating the position where the crossing person existence range is set, and the RAM 103 or the storage 104. To remember. Each detection device 10c is configured to store a crossing person in which the acquired position information is stored when another crossing person existence range and a set of position information are acquired after the crossing person existence range and the set of position information are stored. It is determined whether it is included in the existence range.

  In addition, each detection device 10 c transmits a set of crossing person existence range information and position information stored in the RAM 103 or the storage 104 to the server device 60. The server device 60 stores the set of the crossing person presence range information and the position information transmitted from each detection device 10c in an accumulated manner in the DB 61. Further, the server device 60 transmits a pair of the crossing person existence range information and the position information stored in the DB 61 to each detection device 10c.

  Accordingly, the driver of the vehicle on which each detection device 10c is mounted can know the crossing person existence range at a position where the host vehicle has not traveled in the past, for example. Can pay attention to.

  FIG. 15 is a functional block diagram illustrating an example of the function of the detection device 10c according to the third embodiment. In FIG. 15, the same reference numerals are given to portions common to FIG. 9 described above, and detailed description thereof is omitted.

  In FIG. 15, the detection device 10 c has a storage unit 31, a transmission unit 32, and a reception unit 33 added to the detection device 10 a ′ illustrated in FIG. 9. The storage unit 31 controls writing and reading to and from the RAM 103 and the storage 104 to store and read information. The transmission unit 32 transmits information stored in the storage unit 31 by wireless communication. The receiving unit 33 receives information transmitted by wireless communication and passes it to the storage unit 31.

  In addition to the structure of FIG. 15, the age estimation part 21 and the stop vehicle recognition part 22 which were shown in the above-mentioned FIG. 10 are further added, the age estimated by a crosser, the positional relationship between a crosser and a stop vehicle, etc. Based on the above, the set crossing person presence range may be adjusted.

  FIG. 16 is a functional block diagram of an example for explaining functions of the server device 60 according to the third embodiment. The server device 60 can be configured as a general computer including a CPU, a ROM, a RAM, a storage, a communication I / F, and the like as hardware, and thus detailed description thereof is omitted here. The server device 60 is not limited to being configured as a single device, and the function may be distributed to a plurality of devices, for example.

  In FIG. 16, server device 60 includes a reception unit 62, a transmission unit 63, and a storage unit 64. The storage unit 64 controls writing and reading with respect to the DB 61 to store and read information. The receiving unit 62 receives information transmitted via the network 51 and passes the received information to the storage unit 64. The transmission unit 63 transmits the information stored in the storage unit 64 via the network 51.

  FIG. 17 is a flowchart illustrating an example of processing in the detection apparatus 10c according to the third embodiment. In step S300, the image acquisition unit 11 acquires a captured image. At this time, the image acquisition unit 11 notifies the position information acquisition unit 16 that the captured image has been acquired. In the next step S301, the position information acquisition unit 16 acquires position information indicating the current position in response to the image acquisition notification from the image acquisition unit 11.

  In the next step S302, the setting unit 14 sets the crossing person presence range based on the captured image acquired in step S300. That is, as described with reference to the flowchart of FIG. 4, the detection unit 12 detects the person 203 and the moving area 201 from the captured image acquired by the image acquisition unit 11 (Step S <b> 101 and Step S <b> 102 of FIG. 4). . The calculation unit 13 regards the person 203 as a crossing person when the person 203 detected by the detection unit 12 is present on the moving area 201 (step S103 in FIG. 4), and the captured image and the camera 110. Is calculated from the camera 110 to the crossing person (step S104 in FIG. 4). The setting unit 14 sets the crossing person existence range using the calculated distance.

  In the next step S303, the storage unit 31 associates the position information at the time of imaging acquired in step S301 with the information indicating the crossing person existence range set in step S302, and this position information and the crossing person existence range. And store the first set. Here, the information indicating the crossing person existence range includes the size of the crossing person existence range and information indicating the position of the detected crossing person. The position of the crossing person can be obtained based on the current position at the time of imaging and the distance from the camera 110 to the crossing person.

  In next step S <b> 304, the transmission unit 32 transmits the first set of the position information and the crossing person presence range stored in the storage unit 31 toward the server device 60. Here, every time the first set of the position information and the crossing person presence range is stored in the storage unit 31, the transmission unit 32 transmits the stored first set of the position information and the crossing person existence range. be able to. Further, the transmission unit 32 may collectively transmit the first set of the plurality of position information and the crossing person presence range at regular time intervals or the like.

  The first set of the position information and the crossing person presence range transmitted by the transmission unit 32 is received by the reception unit 62 in the server device 60 via the network 51. The server device 60 stores the first set of the position information and the crossing person presence range received by the receiving unit 62 in the DB 61 by the storage unit 64 as a second set. Further, when the server device 60 receives the first set of the position information and the crossing person existence range from the detection device 10c, the storage unit 64 causes the storage unit 64 to store the one or more pieces of the position information and the crossing person existence range. Read 2 sets. The transmission unit 63 sends the second set of the position information and the crossing person presence range read out by the storage unit 64 to the received detection apparatus 10c of the first set of the positional information and the crossing person existence range. Send. The transmission unit 63 can transmit one or more second sets.

  In step S305 in FIG. 17, the detection device 10c receives the communication from the server device 60, that is, the second set of the position information and the crossing person presence range transmitted from the server device 60, by the reception unit 33. Wait. If there is no communication from the server device 60, the receiving unit 33 returns the process to step S305. If the communication from the server device 60 is received, the receiving unit 33 shifts the process to step S306.

  In step S306, the receiving unit 33 receives the second set of the position information and the crossing person presence range transmitted from the server device 60. In the next step S307, the receiving unit 33 selects the second set of the received position information and the crossing person existence range, and stores the second set of the selected position information and the crossing person existence range in the storage unit. 31 is stored cumulatively. That is, the storage unit 31 includes the first set of the position information acquired and set in the detection device 10c and the crossing person presence range, and the position information transmitted from the server device 60 and received and selected in the detection device 10c. The second set with the crossing person presence range will be stored.

  For example, the receiving unit 33 acquires the current position by the position information acquiring unit 16, and the position information is obtained from the second set of one or more position information and the crossing person presence range received from the server device 60 with respect to the current position. Thus, the second set of the position information within a certain distance and the crossing person presence range can be selected and stored in the storage unit 31. In addition, the receiving unit 33 uses the second set of the one or more pieces of position information and the crossing person existence range received from the server device 60, and the position information and the crossing person existence range where the position information is a certain distance or more with respect to the current position. The second set may not be stored in the storage unit 31. The receiving unit 33 includes a second set of one or more pieces of position information and a crossing person presence range that are not stored in the storage unit 31 among the received two or more pieces of position information and a crossing person existence range. Is discarded, for example.

  Furthermore, the storage unit 31 stores the first set of newly acquired and set position information and the crossing person existence range, and the second set of position information and crossing person existence range transmitted from the server device 60. At the time of storing, a set of a predetermined number of position information and a crossing person existence range that is old in time series is deleted from the first and second sets of the position information and the crossing person existence range that are already stored. May be.

  Furthermore, on the server device 60 side, the second set of the position information and the crossing person existence range is constantly transmitted, and in the detection apparatus 10c, the receiving unit 33 performs the first operation of the transmitted position information and the crossing person existence range. You may make it receive 2 sets for every fixed time. On the server device 60 side, in the second set of the position information stored in the storage unit 64 and the crossing person presence range, the position information is within a certain distance with respect to the position information transmitted from the detection device 10c. The second set of the position information and the crossing person presence range may be selected and transmitted to the detection apparatus 10c.

  When the process of step S307 ends, a series of processes according to the flowchart of FIG. 17 ends.

  The detection apparatus 10c according to the third embodiment includes the position information acquired by the position information acquisition unit 16 in step S301 in the flowchart of FIG. 17, the position information stored in the storage unit 31 in step S307, and the crossing person existence range. And a determination unit that determines whether or not the current position of the host vehicle on which the detection device 10c is mounted is included in any one or more crossing person presence ranges. The determination unit notifies the driver of the own vehicle of the determination result.

  FIG. 18 shows an example of a crossing person presence range displayed on the display 120 of the detection apparatus 10c according to the third embodiment. In the example of FIG. 18, a map 300 including the host vehicle 301 is displayed on the display 120. The map 300 is displayed by the display unit 15 based on, for example, information indicating the current position acquired by the position acquisition device 108 included in the detection device 10c and map information stored in advance in the storage 104, for example.

  In FIG. 18, the detection device 10 c uses the display unit 15 to make a fixed distance from the position of the host vehicle 301 in the first and second sets of the position information stored in the storage unit 31 and the crossing person presence range. The crossing person existence range within is displayed on the map 300. In the example of FIG. 18, the three crosser existence ranges 3101, 311 and 312 are displayed in a display method different from the road on the map 300. In this example, the display colors are different on the map 300. Is displayed. Based on this display, the driver of the host vehicle 301 knows that the certain range of the traveling direction of the host vehicle 301 is, for example, the crossing person existence range 311 before the own vehicle 301 reaches the crossing person existence range 311. Can pay attention in advance to crossers.

  Further, for each crosser existence range 310, 311 and 312, the number of times or the number stored in the storage unit 31 is counted, and the display method of each crosser existence range 310, 311 and 312 is set as the count value of the count result. Further changes may be made accordingly.

  An example of a method for counting the crossing person presence range will be described. For example, when the setting unit 14 sets the crossing person existence range based on the captured image acquired by the image acquisition unit 11, the determination unit described above displays the position information indicating the position where the captured image is acquired as the position information acquisition unit 16. Get from. The determination unit determines whether or not there is a pair including a crossing person existence range including the position indicated by the acquired position information in the pair of the position information and the crossing person existence range stored in the storage unit 31. judge. When the determination unit determines that it exists, the determination unit increments the count value of the set stored in the storage unit 31. The function of the determination unit can be included in the function of the setting unit 14, for example.

  In the example of FIG. 18, for example, a first threshold value and a second threshold value lower than the first threshold value are provided for the count value, and the count values of the crossing person presence ranges 310, 311 and 312 are set. The display colors of the crossing person presence ranges 310, 311 and 312 are changed according to the result of comparison with the first and second threshold values. As an example, the larger the count value, the higher the possibility that a crossing person exists in the crossing person existence range, and a display indicating that the degree of risk is higher.

  Specifically, in FIG. 18, among the crossing person presence ranges 310, 311 and 312, a color (for example, red) indicating that the count value is larger than the first threshold, for example, the crossing person existence range 311 has a high risk level. ), Where the count value is less than the first count value and greater than or equal to the second coefficient value, for example, a color (for example, yellow) indicating that the crossing person presence range 310 has a medium risk level, and the count value is the second threshold value For example, the crossing person presence range 312 is set to a color (for example, green) indicating that the risk level is low.

  In this case, the threshold value is not limited to two, but may be three or more or one. Further, the threshold value may be changed according to the time zone, season, weather conditions, and the like. For example, it is conceivable to change the threshold according to the time zone, and to call the driver more alert at night by setting the night to a threshold lower than the daytime.

  The number of times or the number to be counted is the number or number of crossing person existence ranges set by the detection device 10c mounted on the host vehicle 301 and stored in the storage unit 31, and the storage unit 31 transmitted from the server device 60. The number or number of traverser presence ranges stored in the. However, the number of times or the number of counts may be only the number or number of the crossing person existence ranges set by the detection device 10c mounted on the host vehicle 301 and stored in the storage unit 31.

  As described above, the display methods of the crossing person presence ranges 310, 311 and 312 are made different according to the count value of the number of times or the number of the crossing person existence ranges 310 to 312 stored in the storage unit 31. Thus, it is possible to more clearly indicate the range to be noted to the driver.

  In the above description, the crossing person existence ranges 310 to 312 are displayed on the map 300 based on the position information acquired by the position information acquisition unit 16, but this is not limited to this example. For example, the display unit 15 may display the crossing person existence ranges 310 to 312 overlapping the captured image acquired by the image acquisition unit 11.

(Other embodiments)
The detection program for executing the detection process according to each embodiment is a file in an installable format or an executable format and can be read by a computer such as a CD (Compact Disk) or a DVD (Digital Versatile Disk). Provided by recording on a medium. However, the present invention is not limited to this, and the detection program may be stored in the ROM 102 in advance.

  Furthermore, the detection program for executing the detection process according to each embodiment may be stored on a computer connected to a communication network such as the Internet and provided by being downloaded via the communication network. . Further, a detection program for executing the detection process according to each embodiment and modification may be configured to be provided or distributed via a communication network such as the Internet.

  In the case of the first embodiment, the detection program for executing the detection process according to each embodiment and the modified example is, for example, each of the above-described units (image acquisition unit 11, detection unit 12, calculation unit 13, setting unit 14, and display). The module 101 includes a module 15). As actual hardware, the CPU 101 reads the detection program from, for example, the storage 104 and executes the detection program. Are generated on the main memory.

  In addition, each embodiment is not limited to the above-described one as it is, and can be embodied by modifying constituent elements without departing from the scope of the invention in the implementation stage. Various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the above-described embodiments. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

10, 10a, 10a ′, 10b, 10c Detection device 11 Image acquisition unit 12 Detection unit 13 Calculation unit 14 Setting unit 15 Display unit 16 Position information acquisition unit 21 Age estimation unit 22 Stopped vehicle recognition unit 31 Storage unit 32 Transmission unit 33 Reception Unit 51 Network 60 Server device 61 Database 101 CPU
104 storage 108 position acquisition device 120 display 200 captured image 201 moving area 202 sign 203 person 204 other vehicle 210 crossing person detection position 211a, 211b, 211c crossing person existence range 300 map 301 own vehicle 310, 311, 312 crossing person existence range

Claims (7)

An acquisition unit for acquiring an image captured by the camera;
From the image, at least a detection unit for detecting a person and a region where the moving body can move,
A calculation unit that calculates a first distance from the camera to the person when the person exists in the region;
A detection apparatus comprising: a display unit that displays a range in which the person is estimated to exist according to the position of the person detected by the detection unit and the first distance. The detector is
Detecting a sign on the region from the image;
The detection device according to claim 1, wherein the display unit increases the range as a second distance between the sign and the person becomes shorter. An estimation unit for estimating the age of the person detected by the detection unit;
The detection device according to claim 1, wherein the display unit enlarges the range when the age is not more than a first threshold value. The detection unit further detects a vehicle from the image,
A recognition unit for recognizing a stopped vehicle among the vehicles detected by the detection unit as a stopped vehicle;
The detection device according to claim 1, wherein the range is increased as a third distance between the stopped vehicle and the person becomes shorter. A storage unit for storing a set of the range and position information indicating a position where the image corresponding to the range is acquired;
The detection device according to claim 1, further comprising: a determination unit that determines whether or not a position at which the image used for displaying the range is acquired is included in the range stored in the storage unit. An acquisition step of acquiring an image captured by the camera;
From the image, at least a detection step of detecting a person and a region where the moving body can move;
A calculating step of calculating a first distance from the camera to the person when the person exists in the region;
A detection method comprising: a display step of displaying a range in which the person is estimated to exist according to the position of the person detected by the detection step and the first distance. An acquisition step of acquiring an image captured by the camera;
From the image, at least a detection step of detecting a person and a region where the moving body can move;
A calculating step of calculating a first distance from the camera to the person when the person exists in the region;
A detection program for causing a computer to execute a display step of displaying a range in which the person is estimated to exist according to the position of the person detected by the detection step and the first distance.

Images