JP2010273178A - Stereo measurement system and moving image-reproducing system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a stereo measurement system and a moving image-reproducing system, in which a stored data amount is reduced while securing information necessary for obtaining a situation. <P>SOLUTION: Image data for visual recognition is formed by applying data amount reduction processing to a standard image photographed by a standard camera based on a first frame rate, and three-dimensional state recognition image data is formed from the standard image and a reference image photographed by a reference camera based on a second frame rate, so that the image data for visual recognition and the three-dimensional state recognition image data are stored as stored data. Thus, while securing the information necessary for obtaining the situation, the stereo measurement system can reduce the stored data amount, and the moving image-reproducing system can display three-dimensional information based on the three-dimensional state recognition image data in association with the image data for visual recognition. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a stereo measurement system and a moving image reproduction system, and more particularly to a stereo measurement system and a moving image reproduction system suitable for in-vehicle use.

  In recent years, in the automobile industry, a movement for improving safety has been activated, and introduction of a risk avoidance system using an image sensor, a radar, or the like by a camera is progressing. As a well-known system, there is a system for avoiding danger by acquiring distance information around the host vehicle by using an image sensor, a radar, or the like.

  On the other hand, in the taxi industry, drive recorders are being introduced. A drive recorder is a device that records images before and after an accident, and can be used to analyze the cause of the accident. For example, by viewing a video recorded at the time of a car-to-car collision, it is possible to determine to some extent whether the vehicle is responsible or responsible for other vehicles.

  For example, Patent Document 1 discloses a method capable of recording a long-time image and quickly reproducing a desired image by compressing image information and recording it in a randomly accessible recording unit. ing.

  Furthermore, in Patent Document 2, in an operation management device that uses an image of a drive recorder or the like for training of a driver or the like, if the driving data reaches a dangerous level during the reproduction of an accident image or the like, the image reproduction is slow reproduced. It shows how to make it easier to understand the situation at the time of the accident by switching to.

  In acquiring these images, it is very effective to acquire distance information using a stereo image. For example, in Patent Document 3, a three-dimensional object existing in a monitoring area is extracted using a pair of images captured in a time series by a stereo camera, and a three-dimensional movement of the three-dimensional object is calculated to obtain a monitoring area. A method of detecting a moving object within is shown. However, the purpose of Patent Document 3 is to avoid a collision with a moving object by detecting a moving object in front of the host vehicle, and is mentioned to record an image like the drive recorder described above. Absent.

  On the other hand, in Patent Document 4, in a vehicle traveling state storage device that records an image obtained by an imaging device that captures the periphery of a traveling vehicle, there is an object in a window area provided in the image In addition, there is shown a method of performing distance calculation for each window by stereo measurement, displaying the calculated distance on an image, and storing it together with image information.

Japanese Patent No. 3254946 JP 2008-65361 A JP 2006-134035 A Japanese Patent No. 2608996

  However, in the method of Patent Document 4, what is stored is a composite image of the distance displayed on the image and the image information, and individual information cannot be separated from the stored composite image. Further, if all the composite images of the distance and image information when there is an object in the window area are stored, the amount of data becomes enormous and the storage time is limited or the storage device is required to have a very large capacity. . However, if the number of images to be stored is simply reduced, information is lost and it becomes difficult to grasp the situation in the event of an accident or the like.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a stereo measurement system and a moving image reproduction system in which the amount of data to be saved is reduced while securing information necessary for grasping the situation.

  The object of the present invention can be achieved by the following constitution.

1. A camera unit having at least two cameras: a standard camera that captures a standard image of a stereo image; and a reference camera that captures a reference image of a stereo image;
A recording unit for recording image data captured by the camera unit as saved data;
A control unit for controlling the operation of the camera unit and the recording unit,
In a stereo measurement system that is mounted on a vehicle and acquires a stereo image around the vehicle,
A frame rate determining unit that determines a frame rate of the stored data to be recorded in the recording unit;
A visual image data generation unit that generates visual image data by performing a data amount reduction process on the reference image based on the first frame rate determined by the frame rate determination unit;
A three-dimensional state for generating three-dimensional state grasping image data for grasping a three-dimensional state of a measurement target from the reference image and the reference image based on the second frame rate determined by the frame rate determining unit A grasp image data generation unit,
The recording unit is
The visual image data generated by the visual image data generating unit and the three-dimensional state grasping image data generated by the three-dimensional state grasping image data generating unit are recorded as the stored data. A stereo measurement system.

2. The second frame rate is equal to or lower than the first frame rate;
2. The stereo according to 1, wherein the frame rate determination unit dynamically determines the second frame rate depending on a situation of the vehicle and the surroundings of the vehicle at the time of imaging by the camera unit. Measuring system.

  3. 3. The stereo according to 1 or 2, wherein the frame rate determination unit dynamically determines the second frame rate depending on one or a combination of a plurality of conditions described below. Measuring system.

Condition 1) Speed of the vehicle 2) Operation status of the steering wheel of the vehicle 3) Amount of change in optical flow in at least one of the cameras 4) Time of parallax between the reference camera and the reference camera Amount of change 4. 4. The stereo measurement system according to any one of 1 to 3, wherein the data amount reduction process is a compression process.

  5. 4. The stereo measurement system according to claim 1, wherein the data amount reduction process is an image size reduction process.

  6). 6. The device according to claim 1, wherein the three-dimensional state grasping image data is a set of the reference image and the reference image captured at a timing synchronized with the second frame rate. Stereo measurement system.

  7). 6. The stereo according to claim 1, wherein the three-dimensional state grasping image data is an image including parallax information indicating parallax between the base camera and the reference camera with respect to the measurement target. Measuring system.

  8). The three-dimensional state grasping image data is a distance image including coordinates indicating a position of the measurement target on the visual image data and distance data calculated from the reference image and the reference image. The stereo measurement system according to any one of 1 to 5, characterized in that:

  9. The three-dimensional state grasping image data includes coordinates indicating the position of the measurement target on the visual image data, three-dimensional coordinates calculated from the reference image and the reference image, and measurement target specifying processing. 6. The stereo measurement system according to any one of 1 to 5, wherein the stereo measurement system is a three-dimensional image including a model image that specifies the measurement target based on the three-dimensional image.

10. An image information display section;
A three-dimensional information display unit,
The visual image data recorded by the stereo measurement system according to any one of 1 to 9 is displayed on the image information display unit,
The three-dimensional state recorded by the stereo measurement system according to any one of 1 to 9 in conjunction with the visual image data displayed on the image information display unit on the three-dimensional information display unit A moving image reproduction system characterized by displaying three-dimensional information based on grasp image data.

  According to the present invention, the reference image captured by the reference camera is subjected to the data amount reduction process based on the first frame rate to generate the image data for viewing, and the reference image captured from the reference image and the reference image is captured. Generate 3D state grasping image data based on the second frame rate, save the image data for visual recognition and the 3D state grasping image data as saved data, and secure the information necessary for grasping the situation A stereo measurement system that reduces the amount of data to be performed can be provided, and a moving image reproduction system that displays three-dimensional information based on three-dimensional state grasping image data in conjunction with visual image data can be provided.

It is a block diagram which shows the structure of 1st Embodiment of a stereo measurement system. It is a block diagram which shows the structure of a frame rate determination part. It is a schematic diagram which shows the structure and operation | movement of a visual image data generation part. It is a block diagram which shows the structure of a three-dimensional state grasp image data generation part. It is a schematic diagram which shows the production | generation method of 3D state grasp image data. It is a schematic diagram which shows the production | generation method of the three-dimensional state grasping image data in 2nd Embodiment of a stereo measurement system. It is a schematic diagram which shows embodiment of a moving image reproduction system.

  Hereinafter, the present invention will be described based on the illustrated embodiment, but the present invention is not limited to the embodiment. In the drawings, the same or equivalent parts are denoted by the same reference numerals, and redundant description is omitted.

  First, the configuration of the first embodiment of the stereo measurement system according to the present invention will be described with reference to FIG. 1. FIG. 1 is a block diagram showing the configuration of the first embodiment of the stereo measurement system according to the present invention. FIG.

  In FIG. 1, the stereo measurement system 1 includes a camera unit 11, a recording unit 13, a control unit 15, a sensor unit 17, a data generation unit 19, and the like.

  The camera unit 11 includes at least two cameras such as a reference camera 111 and a reference camera 112. The reference camera 111 and the reference camera 112 are arranged apart from each other by a predetermined baseline length D, and distance information can be acquired from a stereo image. In synchronization with a camera control signal CCS from a camera control unit 151 to be described later, the reference image 111 is output from the reference camera 111 and the reference image Ir is output from the reference camera 112 at a predetermined frame rate FR0.

  The recording unit 13 is composed of, for example, a hard disk, a semiconductor memory, or the like, and in accordance with a recording control signal RCS from a recording control unit 152, which will be described later, visual image data Ds, which will be described later, and three-dimensional state grasping image data Dd are stored data. Record.

  The control unit 15 includes a camera control unit 151, a recording control unit 152, a frame rate determination unit 153, and the like. Each unit of the control unit 13 may be configured by hardware, or the function thereof may be realized using a microcomputer and software.

  The camera control unit 151 outputs a camera control signal CCS for synchronizing the imaging operation of the reference camera 111 and the imaging operation of the reference camera 112.

  The recording control unit 152 controls the recording operation of the recording unit 13 by outputting a recording control signal RCS according to the frame rate determined by the frame rate determining unit 153 described later.

  The frame rate determination unit 153 generates and records visual image data Ds from the reference image Ib captured by the reference camera 111 at a predetermined frame rate FR0 in synchronization with the camera control signal CCS from the camera control unit 151. The first frame rate FR1 to be determined is determined and output to the data generation unit 19.

  Similarly, the frame rate determination unit 153 captures the reference image Ib captured by the reference camera 111 at the predetermined frame rate FR0 in synchronization with the camera control signal CCS from the camera control unit 151, and is captured by the reference camera 112 in the same manner. A second frame rate FR2 for generating and recording the three-dimensional state grasping image data Dd from the reference image Ir thus determined is determined and output to the data generator 19. A method of determining the first frame rate FR1 and the second frame rate FR2 will be described in detail with reference to FIG.

  The sensor unit 17 includes sensors such as a vehicle speed sensor 171 that detects the speed of a vehicle (hereinafter referred to as the host vehicle) on which the stereo measurement system 1 is mounted, and a steering angle sensor 172 that detects an operation state of the steering wheel of the host vehicle. Has been. The vehicle speed signal SS that is the output of the vehicle speed sensor 171 and the steering angle signal HS that is the output of the steering angle sensor 172 are input to the frame rate determination unit 153 and reflected in the determination of the second frame rate FR2. Instead of the steering angle sensor 172, an operation state of the steering wheel of the host vehicle may be detected by detecting acceleration in a direction perpendicular to the traveling direction of the host vehicle using an acceleration sensor.

  The data generation unit 19 includes a visual image data generation unit 191 and a three-dimensional state grasp image data generation unit 192. Each part of the data generation part 19 may be comprised with hardware, and you may implement | achieve the function using a microcomputer and software.

  The visual image data generation unit 191 performs a data amount reduction process such as a compression process or an image size reduction process on the reference thinned image Ib1 obtained by thinning the reference image Ib of the reference camera 111 at the first frame rate FR1. Image data Ds is generated and output to the recording unit 13. Of the reference image Ib input to the visual image data generation unit 191, the image thinned out at the first frame rate FR1 is discarded. Details will be described later with reference to FIG.

  Similarly, the three-dimensional state grasping image data generation unit 192 performs three-dimensional processing on the reference image Ib of the reference camera 111 and the reference image Ir of the reference camera 112 captured in synchronization with the second frame rate FR2. Three-dimensional state grasping image data Dd is generated and output to the recording unit 13. Of the standard image Ib and the reference image Ir input to the three-dimensional state grasping image data generation unit 192, images that are not synchronized with the second frame rate FR2 are discarded. Details will be described later with reference to FIGS.

  Next, a method for determining the first frame rate FR1 and the second frame rate FR2 in the frame rate determination unit 153 in the first embodiment described above will be described with reference to FIG. FIG. 2 is a block diagram showing the configuration of the frame rate determining unit 153.

  In FIG. 2, the frame rate determination unit 153 includes a first frame rate determination unit 1531, a second frame rate determination unit 1532, a parallax change amount calculation unit 1533, an optical flow change amount calculation unit 1534, and the like. Each unit of the frame rate determining unit 153 may be configured by hardware, or the function thereof may be realized using a microcomputer and software.

  The first frame rate determining unit 1531 determines the first frame rate FR1. The first frame rate FR1 is determined to be a predetermined value without depending on the host vehicle and the surroundings of the host vehicle, and is not changed even if changes occur in the host vehicle and the surroundings of the host vehicle. For example, if the reference camera 111 is imaging at a predetermined frame rate FR0 = 30 frames / second (hereinafter referred to as “fps”), the first frame rate FR1 is determined to be 15 fps that is ½ thereof. . As a result, one frame out of two frames of the reference image Ib captured by the reference camera 111 is used for generating the visual image data Ds, and the remaining one frame is discarded.

  On the other hand, the second frame rate determination unit 1532 determines the second frame rate FR2. The second frame rate FR2 is determined at the same rate as the first frame rate FR1 or lower than the first frame rate FR1 depending on the situation of the host vehicle and the surroundings of the host vehicle. As circumstances change, they change dynamically.

  The second frame rate FR2 may be higher than the first frame rate FR1, but will be described later by setting the same rate as the first frame rate FR1 or lower than the first frame rate FR1. In addition, it is possible to reduce the amount of data to be saved while securing information necessary for grasping the situation.

  The second frame rate determination unit 1532 receives a vehicle speed signal SS that is an output of the vehicle speed sensor 171 and a steering angle signal HS that is an output of the steering angle sensor 172 that indicate the state of the host vehicle.

  Further, the base image Ib and the reference image Ir are input to the parallax change amount calculation unit 1533, the parallax change amount calculation unit 1533 calculates the parallax change amount, and the parallax change amount signal Pr is input to the second frame rate determination unit 1532. Entered.

  Similarly, the base image Ib and the reference image Ir are input to the optical flow change amount calculation unit 1534, the optical flow change amount calculation unit 1534 calculates the change amount of the optical flow, and the optical flow change amount signal Of is generated in the second frame. This is input to the rate determination unit 1532. The parallax change amount signal Pr and the optical flow change amount signal Of are signals indicating the situation around the host vehicle.

  The second frame rate determination unit 1532 is based on any one or a combination of a plurality of signals of the vehicle speed signal SS, the steering angle signal HS, the parallax change amount signal Pr, and the optical flow change amount signal Of described above. The second frame rate FR2 is determined and dynamically changed.

  Here, the parallax change amount signal Pr and the optical flow change amount signal Of will be described. First, the parallax change amount signal Pr will be described. The parallax is a shift amount of the position on the screen between the reference image Ir and the reference image Ir where the same subject of interest (hereinafter referred to as the measurement target Os) exists based on the principle of triangulation. The parallax is proportional to the reciprocal of the distance to the measuring object Os, and the larger the parallax is, the closer the distance to the measuring object Os is, and the smaller the parallax is, the farther it is.

  The distance to the measurement target Os can be calculated from the base line length D between the reference camera 111 and the reference camera 112, the focal lengths of the imaging lenses of the reference camera 111 and the reference camera 112, and the parallax value.

  The parallax change amount is a temporal change amount of the above-described parallax. When the parallax change is 0 (zero) or small, there is no change in the distance from the measurement target Os or the change is small. When the parallax changes in the direction in which the parallax increases, the distance to the measurement target Os is short, and when the parallax changes in the direction in which the parallax decreases, the distance from the measurement target Os is long.

  Therefore, when the change in parallax is 0 (zero) or small, or when the parallax changes in the direction in which the parallax decreases, the distance to the measurement target Os, that is, the vehicle, the person in front, the obstacle, or the like does not change or stays away. That is, the possibility of a collision with the measuring object Os is low. On the other hand, when the parallax changes in the increasing direction, the distance to the measurement target Os is approaching, that is, the danger of a collision with the measurement target Os is imminent. Thus, by using the parallax change amount signal Pr, it is possible to know a change in the distance between the host vehicle and the measurement target Os.

  Next, the optical flow change amount signal Of will be described. The optical flow refers to a vector that indicates how the position of the measurement target Os changes in the captured image. When the stereo flow is used to calculate the optical flow of the measurement object Os such as the vehicle ahead, a three-dimensional optical flow can be obtained, and the extension line of the three-dimensional optical flow and the traveling direction of the host vehicle intersect It can be determined that there is a risk of collision.

  Therefore, by using the three-dimensional optical flow, not only the change in the traveling direction of the vehicle such as the distance to the measurement target Os indicated by the above-described amount of parallax change but also the vehicle such as an interruption from the side lane, for example. It is also possible to detect a change in the situation around the vehicle including a change in the direction perpendicular to the traveling direction of the vehicle, and the change in the situation around the vehicle can be more reflected.

  Returning to the second frame rate determination unit 1532, the second frame rate determination unit 1532 is one of the four signals, the vehicle speed signal SS, the steering angle signal HS, the parallax change amount signal Pr, and the optical flow change amount signal Of described above. Based on a combination of one or more signals, the second frame rate FR2 is determined and dynamically changed.

  Here, the host vehicle and the surroundings of the host vehicle are classified into two states, a normal state CS1 and a recording-required state CS2, and the second frame rate FR2 is determined in each state.

(Normal state CS1)
Vehicle speed signal SS: Running at a constant speed or acceleration / deceleration within a predetermined range Steering angle signal HS: Straight ahead or steering angle within a predetermined range Parallax change amount signal Pr: 0 (zero) or smaller, or changing in a direction in which parallax is reduced Optical flow Change amount signal Of: No risk of collision When all the above four conditions are satisfied, it is determined that the risk of collision is low, and the second frame rate FR2 is set to a low rate. For example, if the base camera 111 and the reference camera 112 are imaging at FR0 = 30 fps and the first frame rate FR1 = 15 fps, the second frame rate FR2 is set to ½ that of 7.5 fps. As a result, one of the four frames of the standard image Ib and the reference image Ir is used for generating the three-dimensional state grasping image data Dd, and the remaining three frames are discarded.

(Recording status required CS2)
Vehicle speed signal SS: Acceleration / deceleration exceeding a predetermined range Steering angle signal HS: Steering angle exceeding a predetermined range Parallax change amount signal Pr: Change in the direction of increasing parallax Optical flow change amount signal Of: Risk of collision The above four conditions If one of the above applies, it is determined that there is a risk of collision, and the second frame rate FR2 is set to a high rate. For example, when the base camera 111 and the reference camera 112 are imaging at FR0 = 30 fps and the first frame rate FR1 = 15 fps, the second frame rate FR2 is set to 15 fps, which is the same as the first frame rate FR1. As a result, like the visual image data Ds, one frame is used for generating the three-dimensional state grasping image data Dd for two frames of the standard image Ib and the reference image Ir, and the remaining one frame is discarded.

  It should be noted that it is desirable to use all four signals of the vehicle and the surroundings of the vehicle described above, the vehicle speed signal SS, the steering angle signal HS, the parallax change amount signal Pr, and the optical flow change amount signal Of. It is also possible to use any one or a combination of four signals. For example, if only the vehicle speed signal SS is used and the vehicle speed signal SS indicates constant speed running or acceleration / deceleration within a predetermined range, it is determined as the normal state CS1, and if it indicates acceleration / deceleration exceeding the predetermined range, recording is required. It can also be determined that the state is CS2.

  Next, a method for generating the visual image data Ds in the visual image data generation unit 191 of the data generation unit 19 will be described with reference to FIG. FIG. 3 is a schematic diagram illustrating the configuration and operation of the visual image data generation unit 191. FIG. 3A is a block diagram illustrating the configuration of the visual image data generation unit 191, and FIG. 3B is a visual image. It is a schematic diagram which shows the production | generation process of data Ds.

  In FIG. 3A, the visual image data generation unit 191 includes a thinning unit 1911, a data amount reduction processing unit 1912, and the like. Each unit of the visual image data generation unit 191 may be configured by hardware, or the function thereof may be realized using a microcomputer and software.

  A reference image Ib captured by the reference camera 111 at a predetermined frame rate FR0 (for example, 30 fps) is input to the thinning unit 1911. The thinning unit 1911 generates a reference thinned image Ib1 by thinning the reference image Ib according to the first frame rate FR1 (for example, 15 fps) determined by the frame rate determining unit 153, and outputs it. An image of a frame that has not been used to generate the reference thinned image Ib1 is discarded.

  The reference thinned image Ib1 is subjected to data amount reduction processing by the data amount reduction processing unit 1912 and is output from the visual image data generation unit 191 as visual image data Ds. As the data amount reduction processing, for example, image compression processing, image size reduction processing, or both image compression processing and image size reduction processing can be considered.

  The visual image data Ds only needs to obtain the minimum information necessary for grasping the surrounding situation of the host vehicle at the time of imaging. Therefore, the compression rate or the reduction rate is increased to reduce the data amount of the visual image data Ds. The rate can be increased.

  Next, the generation process of the visual image data Ds will be described with reference to FIG. In FIG. 3B, it is assumed that imaging is progressing in time series along the time axis t shown from the left to the right in the figure. Further, the first frame rate FR1 is set to ½ of the predetermined frame rate FR0. Furthermore, an image indicated by a broken line indicates an image discarded as a result of thinning by the thinning unit 1911.

  In FIG. 3B, the reference image 111 is output from the reference camera 111 at a predetermined frame rate FR0. The reference image Ib is thinned by the thinning unit 1911 at the first frame rate FR1 to generate the reference thinned image Ib1, and the data amount reduction processing unit 1912 performs data amount reduction processing such as compression processing and image size reduction processing. Then, it is recorded in the recording unit 13 as visual image data Ds.

  Next, a method for generating the three-dimensional state grasping image data Dd in the three-dimensional state grasping image data generating unit 192 of the data generation unit 19 will be described with reference to FIGS. 4 and 5. FIG. 4 is a block diagram showing a configuration of the three-dimensional state grasping image data generating unit 192, and FIG. 5 is a schematic diagram showing a method for generating the three-dimensional state grasping image data Dd.

  In FIG. 4, the three-dimensional state grasping image data generation unit 192 includes an image selection unit 1921, a parallax calculation unit 1922, a three-dimensional coordinate calculation unit 1923, a measurement target identification processing unit 1924, and the like. Each unit of the three-dimensional state grasping image data generation unit 192 may be configured by hardware, or the function thereof may be realized using a microcomputer and software.

  A reference image Ib captured at a predetermined frame rate FR0 (for example, 30 fps) by the reference camera 111 and a reference image Ir captured at a predetermined frame rate FR0 by the reference camera 112 are input to the image selection unit 1921.

  The image selection unit 1921 is a standard imaged in synchronization with the second frame rate FR2 (for example, 7.5 fps or 15 fps) determined by the frame rate determination unit 153 among the input standard image Ib and reference image Ir. A set of the image Ib and the reference image Ir is output to the parallax calculation unit 1922 as a set of the selection standard image Ib2 and the selection reference image Ir1. Of the standard image Ib and the reference image Ir input to the image selection unit 1921, images that are not synchronized with the second frame rate FR2 are discarded.

  The parallax calculation unit 1922 uses the set of the input selection reference image Ib2 and the selection reference image Ir1 in the same manner as the parallax change amount calculation unit 1533 of the frame rate determination unit 153 described above, and the vehicle, obstacle, person The parallax signal Ps indicating the distance to the measurement target Os is calculated and output to the three-dimensional coordinate calculation unit 1923. The parallax calculation unit 1922 and the parallax change amount calculation unit 1533 may be shared or partially shared.

  The three-dimensional coordinate calculation unit 1923, based on the input parallax signal Ps, the combination of the selection reference image Ib2 and the selection reference image Ir1, the position coordinates on the visual image data Ds indicating the position where the measurement target Os exists. The three-dimensional coordinate signal Cs indicating the distance to the measurement target Os is calculated and output to the measurement target identification processing unit 1924. Since the visual image data Ds is image data obtained by performing a thinning process and a data amount reduction process on the reference image Ib, the position coordinate of the measurement target Os on the visual image data Ds is the measurement target on the selection reference image Ib2. It can be easily obtained from the position coordinates of Os.

  The measurement target identification processing unit 1924 identifies the measurement target Os based on the three-dimensional coordinate signal Cs indicating the position coordinates and distance of the measurement target Os on the visual image data Ds calculated by the three-dimensional coordinate calculation unit 1923. Then, the measurement target specifying image Ios expressed by modeling the specified measurement target Os is generated, and is output as the three-dimensional state grasping image data Dd together with the combination of the selection reference image Ib2 and the selection reference image Ir1. A specific example of modeling will be described in detail with reference to FIG.

  In the example of FIG. 4, an example in which a set of the measurement target specifying image Ios, the selection reference image Ib2, and the selection reference image Ir1 is output as the three-dimensional state grasping image data Dd is shown. When the processing up to the generation of the measurement target specific image Ios cannot be performed due to the limitation of the processing capability of the data generation unit 192, the processing of each unit of the 3D state grasping image data generation unit 192 described above is performed as much as possible. The output may be output as three-dimensional state grasping image data Dd.

  For example, a set of the selection reference image Ib2 and the selection reference image Ir1 that are the output of the image selection unit 1921 may be output as the three-dimensional state grasping image data Dd, or the parallax signal Ps that is the output of the parallax calculation unit 1922 and The set of the selection standard image Ib2 and the selection reference image Ir1 may be output as the three-dimensional state grasping image data Dd. Also, a set of a three-dimensional coordinate signal Cs indicating the position coordinate and distance of the measurement target Os on the visual image data Ds, which is an output of the three-dimensional coordinate calculation unit 1923, and a selection reference image Ib2 and a selection reference image Ir1. May be output as the three-dimensional state grasping image data Dd.

  Next, a method for generating the three-dimensional state grasping image data Dd will be described with reference to FIG. FIG. 5 shows a process of forming the three-dimensional state grasping image data Dd when returning from the normal state CS1 to the normal state CS1 again through the recording required state CS2.

  In FIG. 5 and FIG. 6 described later, it is assumed that imaging is progressing in time series according to the time axis t shown from the left to the right in the drawing. The first frame rate FR1 is 1/2 of the predetermined frame rate FR0, the second frame rate FR2 is 1/2 of the first frame rate FR1 in the normal state CS1, and the first frame rate in the recording required state CS2. It shall be set equal to FR1. Furthermore, an image indicated by a broken line indicates an image discarded by the image selection unit 1921.

  In FIG. 5, until the timing of time t1, it is the normal state CS1 described above. In this state, the second frame of the low rate in the normal state CS1 in the set of the reference image Ib of the reference camera 111 and the reference image Ir of the reference camera 112 captured at a predetermined frame rate FR0 (for example, 30 fps). A set of images captured in synchronization with the rate FR2 (for example, 7.5 fps) is selected as a set of the selection reference image Ib2 and the selected reference image Ir1.

  The combination of the selected selection reference image Ib2 and the selected reference image Ir1 is subjected to processing in each of the parallax calculation unit 1922, the three-dimensional coordinate calculation unit 1923, and the measurement target identification processing unit 1924, and the three-dimensional state is grasped. Image data Dd is output and recorded at a low second frame rate FR2 in the normal state CS1.

  At the timing of time t1, any one of the four signals, the vehicle speed signal SS, the steering angle signal HS, the parallax change amount signal Pr, and the optical flow change amount signal Of described above, meets the determination condition for the recording state CS2. The state indicating the host vehicle and the surroundings of the host vehicle is switched from the normal state CS1 to the recording required state CS2.

  In this state, the second high-rate signal in the recording-required state CS2 in the set of the reference image Ib of the reference camera 111 and the reference image Ir of the reference camera 112 captured at a predetermined frame rate FR0 (for example, 30 fps). A set of images captured in synchronization with the frame rate FR2 (for example, 15 fps) is selected as a set of the selection reference image Ib2 and the selected reference image Ir1.

  The combination of the selected selection reference image Ib2 and the selected reference image Ir1 is subjected to processing in each of the parallax calculation unit 1922, the three-dimensional coordinate calculation unit 1923, and the measurement target identification processing unit 1924, and the three-dimensional state is grasped. Image data Dd is output and recorded at a high second frame rate FR2 in the recording-required state CS2.

  This state continues until time t2, during which the three-dimensional state grasping image data Dd continues to be recorded at the high second frame rate FR2. Therefore, if a rear-end collision or the like occurs, the visual image data Ds recorded at the first frame rate FR1 and the high-rate second frame rate FR2 in the recording-required state CS2 (here, the first frame rate FR1). And the like, the cause of the accident can be analyzed with higher accuracy.

  Since all the four signals described above have returned to the condition of the normal state CS1 at the timing of the time t2, the state returns to the normal state CS1 again from the state of the required recording state CS2, and the second low-rate second signal in the normal state CS1. The three-dimensional state grasping image data Dd is recorded at the frame rate FR2.

  In all the states that return from the normal state CS1 to the normal state CS1 through the recording-required state CS2 as described above, as shown in FIG. 3B, the visual image data Ds generated from the reference image Ib is recorded. Recording to the unit 13 is also performed.

  As described above, according to the first embodiment, the vehicle speed signal SS, the steering angle signal HS, the parallax change amount signal Pr, and the optical flow change amount signal Of indicating the host vehicle and the surrounding conditions of the host vehicle are displayed. Based on the two signals, the second frame rate FR2 for recording the three-dimensional state grasping image data Dd is dynamically determined, and the three-dimensional state grasping image data Dd is generated at a high frame rate when there is a risk of collision. In this case, it can be connected to a more accurate analysis of the cause of the accident, and in other cases, by generating and recording the three-dimensional state grasping image data Dd at a low frame rate, The recording time can be extended, the capacity of a recording medium such as a hard disk or a semiconductor memory can be reduced, and the apparatus can be reduced in size and cost.

  Next, a second embodiment of the stereo measurement system will be described with reference to FIG. Unlike the first embodiment, the second embodiment does not generate and record the three-dimensional state grasping image data Dd when the vehicle and the surroundings of the vehicle are in the normal state CS1. Only when the recording state CS2 is required, the three-dimensional state grasping image data Dd is generated and recorded in the recording unit 13.

  The formation process of the three-dimensional state grasping image data Dd at this time is shown in FIG. FIG. 6 is a schematic diagram showing a method for generating the three-dimensional state grasping image data Dd in the second embodiment. Similar to FIG. 5, the normal state CS1 returns to the normal state CS1 again through the recording required state CS2. The process of forming the three-dimensional state grasping image data Dd is shown.

  In FIG. 6, until the timing of time t1, the normal state CS1 is described above, the reference camera 111 captures the reference image Ib at a predetermined frame rate FR0, and the reference camera 112 also has the reference image Ir at a predetermined frame rate FR0. However, the generation and recording of the three-dimensional state grasping image data Dd is not performed.

  At the timing of time t1, any one of the four signals, the vehicle speed signal SS, the steering angle signal HS, the parallax change amount signal Pr, and the optical flow change amount signal Of described above, meets the determination condition for the recording state CS2. The state is switched from the normal state CS1 to the recording required state CS2.

  In this state, the second high-rate signal in the recording-required state CS2 in the set of the reference image Ib of the reference camera 111 and the reference image Ir of the reference camera 112 captured at a predetermined frame rate FR0 (for example, 30 fps). A set of images captured in synchronization with the frame rate FR2 (for example, 15 fps) is selected as a set of the selection reference image Ib2 and the selected reference image Ir1.

  The combination of the selected selection reference image Ib2 and the selected reference image Ir1 is subjected to processing in each of the parallax calculation unit 1922, the three-dimensional coordinate calculation unit 1923, and the measurement target identification processing unit 1924, and the three-dimensional state is grasped. Image data Dd is output and recorded at a high second frame rate FR2 in the recording-required state CS2.

  This state continues until time t2, during which the three-dimensional state grasping image data Dd continues to be recorded at the high second frame rate FR2. Therefore, if a rear-end collision or the like occurs, the visual image data Ds recorded at the first frame rate FR1 and the high-rate second frame rate FR2 in the recording-required state CS2 (here, the first frame rate FR1). And the like, the cause of the accident can be analyzed with higher accuracy.

  Since all the four signals described above have returned to the condition of the normal state CS1 at the timing of time t2, the state returns to the normal state CS1 again from the state of the recording required state CS2, and the reference camera 111 and the reference camera 112 capture images. However, the three-dimensional state grasping image data Dd is not generated and recorded.

  In all the states that return from the normal state CS1 to the normal state CS1 through the recording-required state CS2 as described above, as shown in FIG. 3B, the visual image data Ds generated from the reference image Ib is recorded. Recording to the unit 13 is also performed.

  As described above, according to the second embodiment, based on the four signals of the vehicle and the surroundings of the vehicle, the vehicle speed signal, the steering angle signal, the parallax change amount signal, and the optical flow change amount signal. By dynamically determining the second frame rate for recording the three-dimensional state grasping image data, and generating and recording the three-dimensional state grasping image data at a high frame rate only when there is a risk of collision, This can lead to more accurate analysis of the cause of the accident, etc., and if there is no danger, the generation and recording of the 3D state grasping image data is not performed, so the recording time can be extended, and the hard disk, semiconductor memory, etc. Therefore, the capacity of the recording medium can be reduced, and the size and cost of the apparatus can be reduced.

  Next, the moving image reproduction system according to the present invention will be described with reference to FIG. FIG. 7 is a schematic diagram showing the moving image playback system 2 according to the present invention. FIG. 7A shows a first display example, and FIG. 7B shows a second display example.

  7A and 7B, the moving image reproduction system 2 includes an image information display unit 21, a three-dimensional information display unit 23, a reproduction control unit 25, and the like.

  The image information display unit 21 reproduces and displays the visual image data Ds recorded in the recording unit 13 of the stereo measurement system 1 in conjunction with the reproduction control unit 25.

  The three-dimensional information display unit 23 is based on the three-dimensional state grasping image data Dd recorded in the recording unit 13 of the stereo measurement system 1 in conjunction with the visual image data Ds displayed on the image information display unit 21. Replay and display information.

  The reproduction control unit 25 includes a reproduction button 251 and a slider 253. When the reproduction button 251 is pressed, the visual image data Ds is sequentially reproduced and displayed on the image information display unit 21 in accordance with the time axis t in FIG. Further, by sliding the slider 253 to the left and right in the figure, the visual image data Ds for an arbitrary time on the time axis t can be reproduced and displayed on the image information display unit 21.

  In the first display example of FIG. 7A, frames from the imaging timing of the visual image data Ds reproduced and displayed on the image information display unit 21 to a predetermined number of frames out of the three-dimensional information related to the measurement target Os. Only the three-dimensional information based on the three-dimensional state grasping image data Dd is displayed on the three-dimensional information display unit 23. When there is no 3D information based on the 3D state grasping image data Dd in the frames from the imaging timing to a predetermined number of frames, nothing is displayed on the 3D information display unit 23.

  In the second display example of FIG. 7B, all of the three-dimensional information related to the measurement target Os based on the three-dimensional state grasping image data Dd is displayed on the three-dimensional information display unit 23. Then, the three-dimensional information based on the three-dimensional state grasping image data Dd in the frames from the imaging timing of the visual image data Ds reproduced and displayed on the image information display unit 21 to a predetermined number of frames ago models the measurement target Os. The measurement target specifying image Ios (in this case, a rectangular parallelepiped displayed by a broken line) is displayed in a highlighted manner. When the three-dimensional information based on the three-dimensional state grasping image data Dd does not exist in the frames from the imaging timing to a predetermined number of frames, the highlighting display using the measurement target specific image Ios is not performed.

  As described above, according to the moving image reproduction system of the present invention, the three-dimensional information based on the three-dimensional state grasping image data is displayed on the three-dimensional information display unit in conjunction with the visual image data reproduced and displayed on the image information display unit. By playing and displaying the video, even if you use saved data with reduced data volume, the information necessary for grasping the situation is secured, and in the event of a collision, a video that can lead to more accurate analysis of the cause of the accident, etc. A playback system can be provided.

  As described above, according to the present invention, the reference image captured by the reference camera is subjected to the data amount reduction process based on the first frame rate to generate the image data for viewing, and the reference image and the reference camera are used. 3D state grasping image data is generated from the captured reference image based on the second frame rate, and the viewing image data and the 3D state grasping image data are stored as saved data, which is necessary for grasping the situation. Provides a stereo measurement system that reduces the amount of data to be saved while ensuring information, and provides a video playback system that displays 3D information based on 3D state grasping image data in conjunction with visual image data can do.

  It should be noted that the detailed configuration and detailed operation of each component constituting the stereo measurement system and the moving image reproduction system according to the present invention can be changed as appropriate without departing from the spirit of the present invention.

DESCRIPTION OF SYMBOLS 1 Stereo measurement system 11 Camera part 111 Reference camera 112 Reference camera 13 Recording part 15 Control part 151 Camera control part 152 Recording control part 153 Frame rate determination part 1531 1st frame rate determination part 1532 2nd frame rate determination part 1533 Amount of parallax change Calculation unit 1534 Optical flow variation calculation unit 17 Sensor unit 171 Vehicle speed sensor 172 Steering angle sensor 19 Data generation unit 191 Visual image data generation unit 1911 Decimation unit 1912 Data amount reduction processing unit 192 3D state grasp image data generation unit 1921 Image Selection unit 1922 Parallax calculation unit 1923 3D coordinate calculation unit 1924 Measurement target identification processing unit 2 Movie playback system 21 Image information display unit 23 3D information display unit 25 Playback control unit 251 Playback button 253 Slider C S Camera control signal Cs Three-dimensional coordinate signal CS1 Normal state CS2 Required recording state D Baseline length Dd Three-dimensional state grasping image data Ds Visual image data FR0 Predetermined frame rate FR1 First frame rate FR2 Second frame rate HS Steering angle signal Ib reference image Ib1 reference thinned image Ib2 selection reference image Ios measurement target specific image Ir reference image Ir1 selection reference image Of optical flow change amount signal Os measurement target Pr parallax change amount signal Ps parallax signal SS vehicle speed signal

Claims (10)

A camera unit having at least two cameras: a standard camera that captures a standard image of a stereo image; and a reference camera that captures a reference image of a stereo image;
A recording unit for recording image data captured by the camera unit as saved data;
A control unit for controlling the operation of the camera unit and the recording unit,
In a stereo measurement system that is mounted on a vehicle and acquires a stereo image around the vehicle,
A frame rate determining unit that determines a frame rate of the stored data to be recorded in the recording unit;
A visual image data generation unit that generates visual image data by performing a data amount reduction process on the reference image based on the first frame rate determined by the frame rate determination unit;
A three-dimensional state for generating three-dimensional state grasping image data for grasping a three-dimensional state of a measurement target from the reference image and the reference image based on the second frame rate determined by the frame rate determining unit A grasp image data generation unit,
The recording unit is
The visual image data generated by the visual image data generating unit and the three-dimensional state grasping image data generated by the three-dimensional state grasping image data generating unit are recorded as the stored data. A stereo measurement system. The second frame rate is equal to or lower than the first frame rate;
The said frame rate determination part determines the said 2nd frame rate dynamically depending on the condition of the said vehicle at the time of the imaging of the said camera part, and the said vehicle, The said 2nd frame rate is characterized by the above-mentioned. Stereo measurement system. The said frame rate determination part determines the said 2nd frame rate dynamically depending on the combination of any one or several conditions of the following conditions, The Claim 1 or 2 characterized by the above-mentioned. Stereo measurement system.
Condition 1) Speed of the vehicle 2) Operation status of the steering wheel of the vehicle 3) Amount of change in optical flow in at least one of the cameras 4) Time of parallax between the reference camera and the reference camera Amount of change   The stereo measurement system according to claim 1, wherein the data amount reduction process is a compression process.   The stereo measurement system according to any one of claims 1 to 3, wherein the data amount reduction process is an image size reduction process.   6. The device according to claim 1, wherein the three-dimensional state grasping image data is a set of the reference image and the reference image captured at a timing synchronized with the second frame rate. The described stereo measurement system.   6. The three-dimensional state grasping image data is an image including parallax information indicating parallax between the base camera and the reference camera with respect to the measurement target. Stereo measurement system.   The three-dimensional state grasping image data is a distance image including coordinates indicating a position of the measurement target on the visual image data and distance data calculated from the reference image and the reference image. The stereo measurement system according to claim 1, wherein:   The three-dimensional state grasping image data includes coordinates indicating the position of the measurement target on the visual image data, three-dimensional coordinates calculated from the reference image and the reference image, and measurement target specifying processing. The stereo measurement system according to claim 1, wherein the stereo measurement system is a three-dimensional image including a model image in which the measurement target is specified based on the model image. An image information display section;
A three-dimensional information display unit,
The visual image data recorded by the stereo measurement system according to any one of claims 1 to 9 is displayed on the image information display unit,
The three-dimensional information recorded by the stereo measurement system according to any one of claims 1 to 9, in conjunction with the visual image data displayed on the image information display unit, on the three-dimensional information display unit. A moving image reproduction system characterized by displaying three-dimensional information based on state grasping image data.

Images