JP2019028481A - On-vehicle device and driving support apparatus - Google Patents

Abstract

To enable recognition of an important area in any of a plurality of states such as forward, stop, backward or the like of a self-vehicle when detecting a mobile body or the like by processing an image of an on-vehicle device and reduction of a load of image processing.SOLUTION: A range of processing target data in image frames 100A, 100B and 100C is limited by detection frames F11-F17, F21-F25, F31 and F32 and a load of processing is reduced. According to a difference of states such as forward, stop, backward or the like of a self-vehicle, the number of the detection frames, sizes of the detection frames, positions of the detection frames or the like are automatically switched. When travelling forward, the number of the detection frames allocated to processing of a backward video is reduced and the load thereof is reduced, so as to allocate processing capability for that amount to processing of a forward video. Even if a plurality of kinds of driving support functions are simultaneously realized based on the forward image, it is possible to process in realtime without using an expensive CPU.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a vehicle-mounted device and a driving support device having an image processing unit that inputs and processes a video signal output from a vehicle-mounted camera.

  In recent vehicles, an in-vehicle device having an image processing unit for inputting and processing a video signal obtained by photographing with an in-vehicle camera tends to be mounted. For example, a general in-vehicle device that has the function of a drive recorder inputs the video signal of the in-vehicle camera, and when an abnormality such as a traffic accident occurs in the vehicle, the image data within a certain period of time before and after is automatically generated according to the trigger signal. Can be recorded. Some in-vehicle devices having a function of a digital tachograph as an operation recorder input and process a video signal of an in-vehicle camera.

  Further, for example, in the vehicle navigation device disclosed in Patent Document 1, when performing image recognition on the image information around the own position, the recognition processing conditions of the image recognition processing are variably set so that the calculation processing load is increased. Technology to suppress misrecognition and increase the recognition rate. Specifically, a range for performing image recognition processing on each target feature for the image information is set according to the degree of confidence.

  Patent Document 2 shows a vehicle periphery monitoring device that displays an image of a scene around a vehicle by switching between wide area display and enlarged display. The captured wide-area shooting data can be converted into wide-area display data for displaying on the display device with the same field of view and enlarged display data for displaying a part of the wide-area shooting data in an enlarged manner. It is shown that. In addition, when the “front” button is operated while the wide area display data is displayed, an area corresponding to “front” in the wide area shooting data is enlarged and displayed.

JP 2008-298697 A JP 2008-301091 A

By the way, in vehicle-mounted devices, such as a digital tachograph, for example, it is desired to mount various driving support functions such as (1) to (3) shown below.
(1) A function for outputting an alarm when the inter-vehicle distance between the host vehicle and the preceding vehicle is too close.
(2) A function of outputting an alarm when the host vehicle departs from a predetermined travel lane range.
(3) A function for outputting an alarm about an excessive speed of the host vehicle with respect to the speed limit recognized based on the road surface marking.

  When realizing the above functions (1) to (3), for example, a video signal representing a scene in front of the host vehicle photographed by an in-vehicle camera of the host vehicle is processed to perform various pattern recognitions. I need it. That is, based on image data obtained from the video signal, it is necessary to recognize a pattern of a preceding vehicle, a pattern such as a white line representing the left and right boundaries of the traveling lane, a speed limit marking pattern marked on the road surface, and the like.

  In addition, the in-vehicle camera mounted on the vehicle includes an in-vehicle camera that captures a scene behind the host vehicle and an in-vehicle camera that captures an occupant in the vehicle interior in addition to the in-vehicle camera that captures the front of the host vehicle. There is also. Therefore, in the vehicle-mounted device, it may be necessary to simultaneously process a plurality of video signals output from a plurality of vehicle-mounted cameras.

  However, in order to simultaneously recognize in real time a plurality of patterns from images that change sequentially as the host vehicle travels, an expensive hardware such as a high-performance image processing dedicated processor or a high-performance microcomputer is required. Wear must be mounted on the vehicle-mounted device, and there is a concern about an increase in device cost. However, if the above functions are realized with relatively low-cost hardware, a delay occurs in the recognition processing of each pattern, and it becomes difficult to output various alarms for supporting safe driving in real time.

  Therefore, for example, as shown in Patent Document 1 and Patent Document 2, it is assumed that the range to be subjected to image processing in the image is limited in advance. By limiting the range, the load of image processing is reduced, so that the delay of recognition processing can be suppressed to some extent even when relatively low-cost hardware is used.

  However, if a plurality of functions such as the above (1) to (3) are to be realized at the same time, a plurality of recognition objects appearing in different regions in the same video frame must be recognized simultaneously. The range cannot be narrowed. As a result, the amount of data in the recognition target range increases and the delay of the recognition process increases. Also, when processing images from a plurality of in-vehicle cameras, it is necessary to widen the range of the target in consideration of the possibility that the recognition target object exists at various locations in the video frame.

  Also, depending on whether the vehicle is moving forward, stopping, or moving backward, the important areas to be noticed in the video frame are different from each other, so that the important areas can be recognized in any state. , The restriction of the scope of recognition must be relaxed. Accordingly, the amount of data in the recognition target range increases, and the recognition processing delay increases.

  The present invention has been made in view of the above-described circumstances, and an object of the present invention is to enable recognition of an important region in any of a plurality of states such as forward movement, stoppage, and backward movement of the host vehicle, and image processing. It is in providing the onboard equipment and the driving assistance device which can reduce the load of the vehicle.

In order to achieve the above-described object, the vehicle-mounted device and the driving support device according to the present invention are characterized by the following (1) to (5).
(1) An image processing unit that inputs and processes video signals output from one or more in-vehicle cameras mounted on a vehicle;
A vehicle state detector for detecting the presence or absence of travel in the vehicle and at least one state of the traveling direction;
An image region determination unit that dynamically determines a region in the video signal to be processed by the image processing unit;
Have
The image area determination unit automatically changes the area to reflect the detection state of the vehicle state detection unit.
In-vehicle device characterized by that.

  According to the vehicle-mounted device having the configuration (1), it is possible to dynamically determine a region in the video signal processed by the image processing unit in accordance with the detection state of the vehicle state detection unit. Therefore, for example, it is possible to recognize an important area in any of a plurality of states such as forward, stop, and reverse of the host vehicle and optimize the area so that the amount of data to be processed is reduced. The load can be reduced.

(2) The image area determination unit
A constant holding unit for holding constant data representing the position and size of each of a plurality of two-dimensional areas allocated in advance in the video signal;
One or more two-dimensional regions selected according to the detection state of the vehicle state detection unit from among the plurality of two-dimensional regions are allocated as regions in the video signal processed by the image processing unit.
The vehicle-mounted device according to (1) above, wherein

  According to the vehicle-mounted device having the configuration (2), the image area determination unit can easily determine the appropriate position and size of the area processed by the image processing unit based on the constant data held by the constant holding unit. Can be determined.

(3) The image area determination unit automatically determines the number of two-dimensional areas to be selected from the plurality of two-dimensional areas and a combination thereof according to the detection state of the vehicle state detection unit.
The vehicle-mounted device according to (2) above, wherein

  According to the vehicle-mounted device having the configuration of the above (3), for example, a plurality of two-dimensional regions can be assigned simultaneously for each of a plurality of states such as forward, stop, and reverse of the host vehicle. The combination can also be optimized. In addition, even when multiple important points of interest in a video frame are dispersed at positions that are separated from each other, the size of each two-dimensional area is reduced by assigning multiple two-dimensional areas at the same time. In addition, the amount of data processed by the image processing unit can be greatly reduced.

(4) The image area determination unit recognizes one of the forward, backward, and stop states of the vehicle based on the detection state of the vehicle state detection unit, and determines the state according to the recognized vehicle state. Change the area automatically,
The on-vehicle device according to any one of (1) to (3) above,

  According to the vehicle-mounted device having the configuration (4), since any one of the forward, backward, and stopped states of the vehicle can be recognized, an appropriate area corresponding to the state can be assigned as a processing target of the image processing unit. . That is, regardless of whether the vehicle is moving forward, backward, or stopped, it is possible to optimize only the important region of interest in the video frame to be processed and to reduce the amount of data.

(5) The vehicle-mounted device according to any one of (1) to (4) above,
An alarm generating unit that generates an alarm for supporting safe driving of the vehicle based on a result recognized by the image processing unit from the video signal and a predetermined criterion;
A driving support apparatus comprising:

  According to the driving support device having the configuration of (5) above, it is possible to dynamically determine the region in the video signal processed by the image processing unit in accordance with the detection state of the vehicle state detection unit. Therefore, even when realizing a function of simultaneously recognizing a plurality of moving objects from an input video frame and outputting a plurality of types of alarms, the amount of data to be processed can be reduced and alarm delays can be suppressed.

  According to the vehicle-mounted device and the driving support device of the present invention, it is possible to recognize an important region in any of a plurality of states such as forward, stop, and reverse of the host vehicle, and to reduce a load of image processing. Become.

  The present invention has been briefly described above. Further, the details of the present invention will be further clarified by reading through a mode for carrying out the invention described below (hereinafter referred to as “embodiment”) with reference to the accompanying drawings. .

FIG. 1 is a block diagram illustrating a configuration example of an operation management system according to an embodiment of the present invention. FIG. 2 is a flowchart showing an operation example of the digital tachograph for realizing the driving support function. 3 (a), 3 (b), and 3 (c) are front views showing specific examples of the relationship between each image frame obtained in different vehicle states and each detection frame assigned therein. It is. FIG. 4 is a flowchart showing an operation example of a digital tachograph for automatically determining a detection frame to be used according to the state of the vehicle. FIG. 5 is a schematic diagram illustrating a configuration example of the vehicle state table T1 referred to by the digital tachograph. FIG. 6 is a schematic diagram illustrating a configuration example of the detection frame table T2 referred to by the digital tachograph.

  Specific embodiments relating to the present invention will be described below with reference to the drawings.

<Outline of system configuration and operation>
A configuration example of the operation management system 5 in the embodiment of the present invention is shown in FIG.
The operation management system 5 shown in FIG. 1 is introduced as equipment of a business operator such as a trucking company or a bus company. The operation management system 5 manages the operation status of each vehicle such as a truck or a bus. An operation recording device (hereinafter referred to as a digital tachograph) 10 mounted on each vehicle as an in-vehicle device, The office PC 30 is installed in an office or the like. The digital tachograph 10 and the office PC 30 are connected via a network 70 so that they can communicate with each other.

  The office PC 30 is composed of a general-purpose computer device installed in the office, and manages the operation status of the vehicle. The network 70 is a packet communication network such as the Internet to which the radio base station 8 that performs wide area communication with the digital tachograph 10 and the office PC 30 is connected, and relays data communication performed between the digital tachograph 10 and the office PC 30. . Communication between the digital tachograph 10 and the radio base station 8 may be performed by a mobile communication network (portable network) such as LTE (Long Term Evolution) / 4G (4th Generation) or a wireless LAN (Local Area Network). Network).

  The digital tachograph 10 is mounted on a vehicle, and records operation data such as entry / exit time, travel distance, travel time, travel speed, speed over, engine speed over, sudden start, sudden acceleration, sudden deceleration. In addition, the digital tachograph 10 of this embodiment is equipped with a driving support function described later.

  The digital tachograph 10 shown in FIG. 1 includes, as hardware, a CPU (microcomputer) 11, a speed I / F (interface) 12A, an engine rotation I / F 12B, an external input I / F 13, a sensor input I / F 14, and GPS reception. Unit 15, camera I / F 16, nonvolatile memory 26A, volatile memory 26B, recording unit 17, card I / F 18, audio I / F 19, RTC (clock IC) 21, SW input unit 22, communication unit 24, display unit 27, In addition, an analog input I / F 29 is incorporated.

  The CPU 11 comprehensively controls each unit of the digital tachograph 10 according to a program incorporated in advance. The nonvolatile memory 26A holds in advance an operation program executed by the CPU 11, constant data and a table referred to by the CPU 11, and the like. The nonvolatile memory 26A is a rewritable memory, and the stored data can be updated as necessary.

  The recording unit 17 records data such as operation data and video. A memory card 65 possessed by a crew member is detachably connected to the card I / F 18. The CPU 11 writes data such as operation data and video to the memory card 65 connected to the card I / F 18. A built-in speaker 20 is connected to the audio I / F 19. The speaker 20 emits sound such as an alarm.

  The RTC 21 (timer) keeps the current time. The SW input unit 22 is input with ON / OFF signals of various buttons such as a warehousing button and a warehousing button. The display unit 27 is composed of a liquid crystal display (LCD) and displays alarms and the like in addition to communication and operation states.

  A vehicle speed sensor 51 that detects the vehicle speed is connected to the speed I / F 12A, and a speed pulse from the vehicle speed sensor 51 is input. The vehicle speed sensor 51 may be provided as an option in the digital tachograph 10 or may be provided as a device different from the digital tachograph 10. A rotation pulse from an engine rotation speed sensor (not shown) is input to the engine rotation I / F 12B.

  As an input to the external input I / F 13, a brake signal indicating on / off of the brake of the vehicle and a shift signal indicating a shift state of the automatic transmission (including distinction between forward / reverse) are sent from an external device (not shown). Applied.

  The sensor input I / F 14 is connected to an acceleration sensor (G sensor) 28 that detects acceleration (G value) (senses an impact), and receives a signal from the G sensor 28. Examples of the G sensor include those having a method of reading mechanical displacement due to acceleration as vibration or a method of optically reading, but are not particularly limited. The G sensor detects an impact from the front of the vehicle (detects a deceleration G), may also detect an impact from the left and right directions (detects a lateral G), or an impact from the rear of the vehicle. May be detected (acceleration G may be detected). The G sensor is composed of one or a plurality of sensors so that these accelerations can be detected.

  Signals such as a temperature sensor (not shown) for detecting the engine temperature (cooling water temperature) and a fuel amount sensor (not shown) for detecting the fuel amount are input to the analog input I / F 29. CPU11 detects various driving | running states based on the information input via these I / F.

  The GPS receiving unit 15 is connected to the GPS antenna 15a, receives a radio wave of a signal transmitted from a GPS (Global Positioning System) satellite, calculates and acquires information on a current position (GPS position information).

  A plurality of in-vehicle cameras 23 and 23B are connected to the input of the camera I / F 16. In the present embodiment, one on-vehicle camera 23 is installed in the vehicle interior in a state where it is fixed in a direction in which a scene such as a road ahead in the traveling direction of the host vehicle can be photographed. Accordingly, in the video imaged by the in-vehicle camera 23, a preceding vehicle existing ahead of the host vehicle, a white line indicating a traveling lane boundary during traveling, and traffic regulation indications (such as a speed limit) on the road surface appear. The other vehicle-mounted camera 23 is installed in the vehicle compartment in a state where it is fixed in a direction in which a scene such as a road behind the host vehicle can be photographed.

  The in-vehicle cameras 23 and 23B include image sensors in which, for example, 300,000, 1,000,000, and 2,000,000 pixels are arranged on an imaging surface that is imaged through a fisheye lens. The image sensor includes a known sensor such as a CMOS (complementary metal oxide semiconductor) sensor or a CCD (charge coupled device) sensor.

  The video signals output from the in-vehicle cameras 23 and 23B are converted into digital signals representing gradation and color for each pixel by the internal circuit of the camera I / F 16 and output from the camera I / F 16 as image data for each frame. The

  The video (image data) taken by each of the in-vehicle cameras 23 and 23B is recorded as time series data by the operation of the recording unit 17, but is repeatedly overwritten so as to be recorded for a predetermined time. The predetermined time is a time corresponding to several seconds (for example, 2 seconds, 4 seconds, 10 seconds, etc.) before and after the accident so that the situation of the accident can be understood when the accident occurs. The image captured by the camera 23 may be a still image or a moving image. Images before and after the accident are displayed on the display unit 33 of the office PC 30.

Moreover, the digital tachograph 10 of this embodiment is equipped with driving support functions such as (1) to (4) shown below.
(1) A function for outputting an alarm when the distance between the host vehicle and the preceding vehicle is too close.
(2) A function of outputting an alarm when the vehicle deviates from the range of the traveling lane in which the host vehicle is traveling.
(3) A function of outputting an overspeed warning when the traveling speed of the host vehicle exceeds the speed limit of the road marking.
(4) A function that informs the driver of the presence of surrounding obstacles in situations such as when the host vehicle is moving backward.

  In order to realize the functions (1) to (4) described above, it is necessary to perform predetermined image recognition by processing image data of videos taken by the in-vehicle cameras 23 and 23B. In other words, the position and distance of the preceding vehicle are identified by image recognition, the relative position between the white line of the driving lane boundary and the host vehicle is detected, the speed limit of the road marking is detected, and various obstacles are detected. It will be necessary.

  Image recognition as described above can be realized by executing a predetermined recognition algorithm according to a program in which the CPU 11 is incorporated. However, when the amount of data of the image to be processed is large, the load on the CPU 11 required for image recognition becomes very large, so that processing in real time becomes difficult and detection delay may occur. In particular, when images from a plurality of in-vehicle cameras 23 and 23B are processed simultaneously or a plurality of recognition targets are detected at the same time, there is a concern about the occurrence of delay. Therefore, in this embodiment, when executing image recognition, a detection frame is provided as described later to limit the data range to be processed, and special processing is performed according to the state of the vehicle.

  When the communication unit 24 performs wide area communication and is connected to the wireless base station 8 via a mobile network (mobile communication network), the communication unit 24 communicates with the office PC 30 via the network 70 such as the Internet connected to the wireless base station 8. Communicate. The power supply unit 25 supplies power to each unit of the digital tachograph 10 when the ignition switch is turned on.

  On the other hand, the office PC 30 is configured by a PC (personal computer) that operates on a general-purpose operating system. The office PC 30 functions as an operation management device and includes a CPU 31, a communication unit 32, a display unit 33, a storage unit 34, a card I / F 35, an operation unit 36, an output unit 37, a voice I / F 38, and an external I / F 48. .

  The CPU 31 comprehensively controls each unit of the office PC 30. The communication unit 32 can communicate with the digital tachograph 10 via the network 70. Further, the communication unit 32 can be connected to various databases (not shown) connected to the network 70 and can acquire necessary data.

  In addition to the operation management screen, the display unit 33 displays an accident video, a hazard map, and the like. The storage unit 34 holds various programs such as system analysis software for displaying video received from the digital tachograph 10 and displaying vehicle position information on a map.

  A memory card 65 is detachably attached to the card I / F 35. The card I / F 35 receives operation data measured by the digital tachograph 10 and stored in the memory card 65. The operation unit 36 includes a keyboard, a mouse, and the like, and accepts operations of an office manager. The output unit 37 outputs various data. A microphone 41 and a speaker 42 are connected to the audio I / F 38. The manager of the office can also make a voice call using the microphone 41 and the speaker 42. When a vehicle accident occurs, the manager of the office makes an emergency call or contacts the police.

  An external storage device (storage memory) 54 is connected to the external I / F 48. The external storage device 54 stores an accident point database (DB) 55, an operation data DB 56, and a hazard map DB 57. In the accident point database (DB) 55, GPS position information (latitude and longitude) of the vehicle at the time of the accident transmitted from the digital tachograph 10 is registered. In the operation data DB 56, as the operation data, in addition to loading / unloading time, speed, travel distance, etc., rapid acceleration / deceleration, sudden steering, speed over, engine speed over, etc. are recorded. Registered in the hazard map DB 57 is map data in which marks representing points where accidents have occurred in the past (accident points) are superimposed on the map. In addition, the hazard map may describe an area where a disaster such as a natural disaster is expected, an evacuation site, or the like.

  When the CPU 31 reads out the hazard map of the designated area (for example, the area including the accident point) from the hazard map DB 57 and displays it on the display unit 33, the CPU 31 acquires the data of the accident point registered in the accident point DB 55, Marks representing these accident points are superimposed on the map to generate a new hazard map. The manager of the office can immediately see the latest accident location on the map (hazard map).

<Details of driving support function>
An example of the operation of the digital tachograph 10 for realizing the driving support function is shown in FIG. The operation shown in FIG. 2 is repeatedly executed in a short time period by the CPU 11 in the digital tachograph 10.

  In step S21, the CPU 11 executes a predetermined image recognition process on the image data of each frame corresponding to the video captured by each of the in-vehicle cameras 23 and 23B. However, the data to be processed is limited only to the range of each detection frame defined in advance in each image frame. The range of each detection frame can be specified by referring to the detection frame table T2A, for example. The detection frame table T2A will be described later.

  In step S22, the CPU 11 executes the inter-vehicle distance detection process using the recognition result in S21. That is, the positional relationship between the position of the preceding vehicle detected in the image frame and the own vehicle is grasped, and the inter-vehicle distance is calculated based on the distance on the image and the photographing condition of each camera. Further, the calculated inter-vehicle distance is compared with a predetermined threshold, and an inter-vehicle distance warning is output when the inter-vehicle distance is too close.

  In step S23, the CPU 11 executes a lane departure detection process using the recognition result in S21. That is, by calculating the positional relationship and distance in the left-right direction between the position of the white line on the lane boundary detected in the image frame and the position of the host vehicle, and comparing the result with a predetermined determination condition, Identify any deviations. If there is a lane departure, a lane departure warning is output.

  In step S24, the CPU 11 executes the overspeed detection process using the recognition result in S21. That is, the speed limit recognized based on the speed limit indication on the road surface is compared with the current traveling speed of the host vehicle to identify the presence or absence of excessive speed. If there is an overspeed, an overspeed warning is output.

  In step S25, the CPU 11 performs detection processing other than the above using the recognition result in S21. For example, the presence / absence of various obstacles, the presence / absence of a moving object, and various boundary positions on the road are detected. Then, an alarm is output when an approach to an obstacle or a moving object is detected, or when the host vehicle approaches the boundary position.

<Relationship between image frame and detection frame>
Specific examples of the relationship between the image frames 100A, 100B, and 100C obtained in different vehicle states and the detection frames allocated therein are shown in FIGS. 3A, 3B, and 3C. Respectively.

  When the CPU 11 recognizes the image data obtained by the in-vehicle cameras 23 and 23B, if the amount of image data is large, the processing load on the CPU 11 increases, and real-time processing becomes difficult and delay occurs. . Therefore, in order to reduce the processing load on the CPU 11, the amount of data to be processed is reduced. Specifically, a detection frame that defines the range of data to be processed is provided.

  In particular, when processing the image of the in-vehicle camera 23B that captures the rear, the detection frame is optimized so as to be optimized according to whether the vehicle state is “reverse”, “stop”, or “forward”. The total number of frames, the position of each detection frame, the size of each detection frame, and the like are changed. In other words, depending on whether the vehicle status is “reverse”, “stop”, or “forward”, there are different important points to be noted in the image frame. Thus, each detection frame can be minimized, and the amount of data to be processed can be effectively reduced.

  3 (a), 3 (b), and 3 (c) are assigned to each image frame 100A, 100B, and 100C in the “reverse”, “stop”, and “forward” states, respectively. The example with the relationship with each detection frame is shown.

  In the image frame 100A shown in FIG. 3A, seven mutually independent rectangular detection frames F11, F12, F13, F14, F15, F16, and F17 are assigned to different positions in the frame. In the image frame 100B shown in FIG. 3B, five mutually independent rectangular detection frames F21, F22, F23, F24, and F25 are assigned to different positions in the frame. In the image frame 100C shown in FIG. 3C, two mutually independent rectangular detection frames F31 and F32 are assigned to different positions in the frame.

  That is, when the host vehicle moves backward, as in the detection frames F11 to F17 shown in FIG. 3A, attention is paid to various points in the rear image, that is, the rear periphery of the host vehicle and the destination in the traveling direction. Therefore, sufficient safety can be ensured during retreat. When the host vehicle is stopped, as shown in the detection frames F21 to F25 shown in FIG. 3B, the host vehicle stops by paying attention to the vicinity of the host vehicle at a short distance in the rear image. Sufficient safety can be secured when starting to move from the state. Further, when the host vehicle is moving forward, it is necessary to pay attention to only the right and left sides at a short distance in the rear image as in the detection frames F31 and F32 shown in FIG. Can be secured.

  On the other hand, when the host vehicle is moving forward, it is indispensable to recognize the front image captured by the in-vehicle camera 23 at the same time as the rear image captured by the in-vehicle camera 23B, and the processing load of the CPU 11 increases. Here, if the front image and the rear image are processed simultaneously, the load on the CPU 11 further increases. However, by limiting the target range of image processing to only the detection frames F31 and F32 shown in FIG. 3C in the rear video, the amount of data to be processed can be reduced and the increase in load can be suppressed.

<Automatic detection frame switching according to vehicle conditions>
FIG. 4 shows an operation example of the digital tachograph 10 for automatically determining a detection frame to be used according to the state of the vehicle. That is, when the CPU 11 of the digital tachograph 10 executes the operation shown in FIG. 4, the detection frames are automatically switched as shown in FIGS. 3A, 3 </ b> B, and 3 </ b> C, for example. Realize. The operation of FIG. 4 will be described below.

  In step S11, the CPU 11 identifies forward / reverse of the host vehicle. For example, by referring to the state of the shift signal obtained from the automatic transmission of the host vehicle, it is possible to distinguish whether the host vehicle is moving forward or backward.

  In step S12, the CPU 11 identifies whether or not the host vehicle is stopped. For example, the presence or absence of a stop can be distinguished based on information on the current traveling speed of the host vehicle and the state of the brake signal.

  In step S13, the CPU 11 distinguishes whether the host vehicle is in the “reverse”, “stop”, or “forward” state based on the results of S11 and S12. In the case of “reverse”, the process proceeds from S13 to S14. In the case of “stop”, the process proceeds from S13 to S15. In the case of “forward”, the process proceeds from S13 to S16.

  In step S14, the CPU 11 obtains constant data representing the detection frame combination at the time of reverse from the vehicle state table T1 and the detection frame table T2. The vehicle state table T1 and the detection frame table T2 will be described later.

  In step S15, CPU11 acquires the constant data showing the detection frame combination at the time of a stop from vehicle state table T1 and detection frame table T2. In step S16, the CPU 11 acquires constant data representing the detection frame combination at the time of forward movement from the vehicle state table T1 and the detection frame table T2.

  In step S17, the CPU 11 writes the detection frame combination constant data acquired in any of S14, S15, and S16 in the detection frame table T2A. The contents of the detection frame table T2A are reflected in the detection frame used in the process of step S21 shown in FIG. Note that the detection frame table T2A may be omitted and a part of the detection frame table T2 may be selectively used in the process of S21.

<Configuration example of tables T1 and T2>
A configuration example of the vehicle state table T1 referred to by the digital tachograph is shown in FIG. Further, FIG. 6 shows a configuration example of the detection frame table T2 referred to by the digital tachograph.

  The vehicle state table T1 shown in FIG. 5 holds constant data representing the relationship between the states of “reverse”, “stop”, and “forward”, the number of detection frames N1, and the number N2 at the start of the detection frame. ing. In the vehicle state table T1 of FIG. 5, the number N1 of detection frames at the time of reverse movement is “7” as in the example shown in FIG. Further, as in the example shown in FIG. 3B, the number of detection frames N1 when the vehicle is stopped is “5”. Further, as in the example shown in FIG. 3C, the number N1 of detection frames during forward movement is “2”.

  The detection frame table T2 shown in FIG. 6 holds constant data representing the relationship among the number N3, the reference position P, and the detection frame size S for each of a large number of detection frames. In this detection frame table T2, for example, a detection frame with the number N3 “1” is a constant indicating that the reference position P is P1, the size in the X-axis direction is Sx1, and the size in the Y-axis direction is Sy1. Data is allocated.

  For example, when the number N21 at the top of the detection frame in the vehicle state table T1 shown in FIG. 5 is “1”, the constant data within the range where the number N3 in the detection frame table T2 is “1” to “7” The combination of the detection frames at the time of reverse is selected in S14 of FIG. Therefore, it is possible to specify the position in the frame and the vertical / horizontal sizes of each of the seven detection frames F11 to F17 shown in FIG.

  Further, when the number N22 at the top of the detection frame in the vehicle state table T1 shown in FIG. 5 is “8”, the constant data within the range where the number N3 in the detection frame table T2 is “8” to “12” The combination of detection frames when the vehicle is stopped is selected in S15 of FIG. Therefore, it is possible to specify the position and the vertical / horizontal sizes in each of the five detection frames F21 to F25 shown in FIG.

  Further, when the number N23 at the top of the detection frame in the vehicle state table T1 shown in FIG. 5 is “13”, the constant data within the range where the number N3 in the detection frame table T2 is “13” to “14” The combination of detection frames at the time of forward movement is selected in S16 of FIG. Therefore, it is possible to specify the position and the vertical / horizontal size in each of the two detection frames F31 to F32 shown in FIG.

  A part of the data extracted in any one of S14, S15, and S16 from the detection frame table T2 is written to the detection frame table T2A, and the contents of the detection frame table T2A are reflected in the process of step S21 in FIG. . Therefore, depending on whether the state of the host vehicle is “reverse”, “stop”, or “forward”, as shown in FIGS. 3 (a), 3 (b), and 3 (c) It is possible to automatically switch the number, position, size, etc. of the detection frames used for.

  The vehicle state table T1 and the detection frame table T2 are arranged in a storage area on the nonvolatile memory 26A, and each constant data held by these is changed as necessary. For example, when the in-vehicle camera 23B is newly installed in the host vehicle or the position thereof is changed, special software for adjusting each parameter of the digital tachograph 10 is executed by the CPU 11 so that the manager can change the vehicle state table. T1 and the detection frame table T2 can be updated.

<Possibility of deformation>
In the example shown in FIG. 3A, FIG. 3B, and FIG. 3C, only the detection frame in the case of processing the video of the in-vehicle camera 23B that captures the rear is assumed, but the front is captured. In the case of processing the video of the in-vehicle camera 23 or other in-vehicle camera that captures the side of the vehicle or the interior of the vehicle, it is assumed that a similar detection frame is applied to reduce the amount of data to be processed. . For example, when the host vehicle is moving backward, the load on the CPU 11 is reduced by reducing the number and size of the detection frames used by the process for recognizing the front image, and the delay in the process for recognizing the rear image. Can be suppressed.

<Advantages of digital tachograph 10>
In the digital tachograph 10 shown in FIG. 1, when the operation shown in FIG. 2 is executed to realize the driving support function, for example, when the host vehicle is moving forward, the detection frame is used in step S21. Thus, it is possible to greatly reduce the load on the CPU 11 for processing the rear video. As a result, even when trying to simultaneously realize multiple types of driving support functions (such as inter-vehicle distance warning, lane departure warning, overspeed warning, etc.) based on the video taken in front, a high-performance CPU Real time processing without delay becomes possible without using. Further, for example, when the host vehicle moves backward, the number of detection frames used when processing the rear image can be increased in S14 as shown in FIG. 3A, so that the rear safety can be sufficiently ensured. .

Here, the features of the vehicle-mounted device and the driving support device according to the embodiment of the present invention described above are briefly summarized and listed in the following [1] to [5], respectively.
[1] An image processing unit (CPU11, S21) that inputs and processes video signals output from one or more in-vehicle cameras (23, 23B) mounted on the vehicle;
A vehicle state detection unit (CPU 11, S11, S12) for detecting the presence or absence of travel in the vehicle and the state of at least one of the traveling directions;
An image area determination unit (CPU 11, S13 to S16) for dynamically determining an area in the video signal processed by the image processing unit;
Have
The image region determination unit automatically changes the region to reflect the detection state of the vehicle state detection unit (S17, S21),
In-vehicle device characterized by that.

[2] The image area determination unit
A constant holding unit (detection frame table) that holds constant data representing the position and size of a plurality of two-dimensional areas (detection frames F11 to F17, F21 to F25, F31, and F32) allocated in advance in the video signal. T2)
One or more two-dimensional regions selected from the plurality of two-dimensional regions according to the detection state of the vehicle state detection unit are assigned as regions in the video signal processed by the image processing unit (S17). ,
The on-vehicle device according to [1] above, wherein

[3] The image region determination unit automatically determines the number of two-dimensional regions to be selected from the plurality of two-dimensional regions and a combination thereof according to the detection state of the vehicle state detection unit (S13 to S13). S16),
The on-vehicle device according to [2] above,

[4] The image area determination unit recognizes one of the forward, backward, and stop states of the vehicle based on the detection state of the vehicle state detection unit, and determines the state according to the recognized vehicle state. The area is automatically changed (S13 to S16),
The vehicle-mounted device according to any one of [1] to [3], wherein

[5] The vehicle-mounted device according to any one of [1] to [4],
An alarm generation unit (S22 to S25) for generating an alarm for supporting safe driving of the vehicle based on a result recognized by the image processing unit from the video signal and a predetermined determination criterion;
A driving support apparatus comprising:

5 Operation management system 8 Wireless base station 10 Digital tachograph 11, 31 CPU
12A Speed I / F
12B Engine rotation I / F
13 External input I / F
14 Sensor input I / F
15 GPS receiver 15a GPS antenna 16 Camera I / F
17 Recording unit 18 Card I / F
19 Voice I / F
20, 42 Speaker 21 RTC
22 SW input unit 23 In-vehicle camera 24, 32 Communication unit 25 Power supply unit 26A Non-volatile memory 26B Volatile memory 27 Display unit 28 G sensor 29 Analog input I / F
30 office PC
33 Display unit 34 Storage unit 35 Card I / F
36 Operation unit 37 Output unit 38 Audio I / F
41 Microphone 48 External I / F
51 Vehicle speed sensor 54 External storage device 55 Accident point DB
56 Operation data DB
57 Hazard Map DB
65 Memory card 70 Network 100A, 100B, 100C Image frame F11, F12, F13, F14, F15, F16, F17 Detection frame F21, F22, F23, F24, F25, F31, F32 Detection frame T1 Vehicle state table T2, T2A Detection Frame table

Claims (5)

An image processing unit that inputs and processes video signals output from one or more in-vehicle cameras mounted on the vehicle;
A vehicle state detector for detecting the presence or absence of travel in the vehicle and at least one state of the traveling direction;
An image region determination unit that dynamically determines a region in the video signal to be processed by the image processing unit;
Have
The image area determination unit automatically changes the area to reflect the detection state of the vehicle state detection unit.
In-vehicle device characterized by that. The image area determination unit
A constant holding unit for holding constant data representing the position and size of each of a plurality of two-dimensional areas allocated in advance in the video signal;
One or more two-dimensional regions selected according to the detection state of the vehicle state detection unit from among the plurality of two-dimensional regions are allocated as regions in the video signal processed by the image processing unit.
The on-vehicle device according to claim 1. The image region determination unit automatically determines the number of two-dimensional regions to be selected from the plurality of two-dimensional regions and a combination thereof according to the detection state of the vehicle state detection unit.
The vehicle-mounted device according to claim 2. The image region determination unit recognizes one of the forward, backward, and stopped states of the vehicle based on the detection state of the vehicle state detection unit, and automatically determines the region according to the recognized vehicle state. Change
The on-vehicle device according to any one of claims 1 to 3, wherein The vehicle-mounted device according to any one of claims 1 to 4,
An alarm generating unit that generates an alarm for supporting safe driving of the vehicle based on a result recognized by the image processing unit from the video signal and a predetermined criterion;
A driving support apparatus comprising:

Images