JPH0664074B2 - Moving object speed measuring device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a velocity measuring device for a moving object, and particularly to a movement suitable for obtaining an unsteady three-dimensional velocity vector of a plurality of moving objects of arbitrary shapes. The present invention relates to a speed measuring device for an object.

[Conventional technology]

As a conventional example, there is a device shown in Japanese Patent Application No. 59-30281 "Velocity distribution measuring device for moving object". In this conventional example, an imaging device that captures a moving object, an A / D converter that A / D converts a signal from the imaging device, a superimposing circuit that superimposes a signal from the A / D converter for each frame, and a superimposing device. A velocity measuring device comprising a velocity calculating circuit for obtaining a velocity of a moving object from superimposition data from the circuit, wherein a background removing circuit is provided between the A / D converter and the superimposing circuit and the A / D is set at a predetermined time. A distributor that sends the signal from the converter to the velocity calculator for each frame is provided, and the image of the same moving object in the superimposed data is identified using the moving object image in the signal from the distributor. It is characterized in that a speed calculator capable of obtaining the moving direction is provided. Further, as a background removal method, a signal sequence having a difference between two consecutive frames is sequentially superimposed. This conventional example has one frame memory before the speed calculator,
The components of the device are arranged in series.

[Problems to be solved by the invention]

In the above-mentioned conventional example, the frame memory provided before the speed calculator is used for background removal. In this configuration, the following two background removal methods can be implemented. (1)
The image of the measurement start point including the background is stored in the frame memory, the difference from the input video is taken using the inverter and the adder, and the superimposition processing is performed to remove only the background, which is the static component, and Leave only the image. (2) Two consecutive background removals
This is performed between the signal trains that take the difference between frames. In this case, the contents of the frame memory are replaced every time.

In the method (1), in the trajectory image, when the image of the starting point of the moving object and the image of the moving object in the next video have the same gray level in the overlapping part, even though the image is not a still image, Eighth, erased by the background removal function
As shown in the figure, the locus image may be cut at the position of the image of the starting point of the moving object, and it may be difficult to identify the image of the starting point of the moving object and recognize the traveling direction of the moving object. There's a problem.

In the method (2), since the background is removed between two consecutive frames, the overlapping portions of the images of the moving object between the two consecutive frames are erased as in the case of the method (1). As shown in FIG. 9, there is a problem that the trajectory image becomes fragmented, and it becomes difficult to identify the end point of the moving object from the image or the like and measure the length.

Further, in this conventional technique, since the respective constituent elements of the device are arranged in series and the calculation between images cannot be freely selected and executed, the measurement start point necessary for recognizing the traveling direction of the moving object Background removal cannot be applied to images such as measurement end points. Therefore, there is a problem that the background image remains as shown in FIGS. 10 and 11, and the background interferes with the identification of the trajectory image.

Further, in this conventional technique, when the speed calculator cannot obtain the speed and moving direction of the moving object in real time, it is possible to send the superposition data and the data of the measurement start point and the measurement end point to the speed calculator. This is not possible, and it becomes difficult to obtain the time-series image data required for obtaining the non-steady-moving velocity of the non-steady-moving object at high speed.

An object of the present invention is to easily specify the trajectory of each moving object even when there are a plurality of moving objects, and to efficiently obtain the moving speed of each moving object.

[Means for solving problems]

A feature of the present invention that achieves the above object is a means for obtaining moving object start point image data at a starting point of movement of a moving object during photographing by a photographing device, and a movement at an end point of movement of the moving object during the photographing. Means for obtaining object end point image data, a first memory for storing the moving object start point image data, a second memory for storing the moving object end point image data, and the movement based on the video signal with the background removed. Trajectory data specified by specifying a trajectory of the moving object whose velocity is to be determined by using trajectory creating means for determining the trajectory of the object, the moving object start point image data, the moving object end point image data, and the trajectory of the moving object. And means for obtaining the speed of the moving object based on the above.

[Action]

Since the trajectory of the moving object whose speed is to be determined is specified using the moving object start point image data, the moving object end point image data, and the trajectory of the moving object, the trajectory of the moving object that determines the moving speed even when there are multiple moving object trajectories. Can be obtained easily and accurately. In particular, since the trajectory of the corresponding moving object can be specified by using the moving object starting point image data and the moving object end point image data, even when the trajectories of a plurality of moving objects intersect, the corresponding trajectory can be easily specified. It Since the moving speed of the corresponding moving object is obtained using the specified trajectory data, the moving speed can be obtained efficiently.

〔Example〕

Hereinafter, a moving object velocity measuring apparatus according to a preferred embodiment of the present invention will be described with reference to FIGS. 1 to 3. The moving object velocity measuring apparatus of this embodiment includes a TV camera 1 for capturing an image of a moving object, a video input circuit 14 for inputting a video signal output from the TV camera 1, a trajectory creating circuit 2, a trajectory image memory 3, and a background. The image memory 4, the start point image memory 5, the end point image memory 6, the communication control circuit 13, the image analysis device 10,
The monitor 16 is composed of a video output circuit 15 for outputting the created image as a video. The locus creation circuit 2 has a binarization circuit 7, a superposition processing circuit 8 and a background removal circuit 9. Then, the video input circuit 14, the video output circuit 15, the binarization circuit 7, the superposition processing circuit 8, the background removal circuit 9, the trajectory image memory 3, the background image memory 4, the start point image memory 5, the end point image memory 6, and the communication control circuit. Reference numerals 13 are respectively connected in parallel to buses 12 connected to an arithmetic processing circuit (CPU) 11. Transmission control of video signals and image signals via the bus 12 is performed by the CPU 11. Also,
It has a display device 40, a printer 41, a plotter 42, and a disk device 43 for outputting or recording the result of image analysis.

Next, the operation of this embodiment will be described. That is, the video signal output from the TV camera 1 is processed by the following procedure,
Create a trajectory image of a moving object. (A) A video image of the background before the moving object is captured using the TV camera 1 is previously captured, and the video image of the background is input and stored in the background image memory 4. (B) The TV camera 1 is used to capture an image of a moving object. (C) The trajectory creating circuit 2 removes the background image stored in the background image memory 4 from the moving object video signal, and superimposes the background image stored on the trajectory image memory 3 to generate a new trajectory image signal of the moving object. It is output to the trajectory image memory 3. The processing of (b) and (c) is performed during a predetermined measurement period,
The trajectory image signal of the moving object to which the trajectory of the moving object is sequentially added is stored in the trajectory image memory 3 by repeating at a predetermined time interval. At this time, the CPU 11 stores the video signal of the first moving object after the start of measurement in the video of the moving object in the designated measurement period as the start point video signal in the start point image memory 5, and at the end of the measurement time. End point image memory 9 using the video signal of the moving object as the end point image signal
To memorize. The video signal captured by the TV camera 1 and input from the video input circuit 14 is once input into the CPU 11. The CPU 11 selects a start-point image signal and an end-point image signal and inputs them to the respective image memories 5 and 6. The CPU 11 also inputs the video signal during the measurement period to the background removal circuit 9.

The CPU 11 causes the background image memory 4 to input the video signal of the background before the moving object is photographed. The CPU 11 also controls signal transmission among the memories connected to the bus 12, the trajectory creation circuit 2, the communication control circuit 13, and the video output circuit 15.

Next, a procedure for creating a trajectory image signal of a moving object will be described in more detail with reference to FIG. Of the video signal 47 output from the TV camera 1 and input to the bus 12 via the video input circuit 14, the background image captured before capturing the moving object is first stored in the background image memory 4. The moving object video signal 47 output from the TV camera 1 and the background video signal stored in the background image memory 4 are input to the background removal circuit 9 in response to an instruction to start moving object measurement. That is, the background video signal is bit-inverted by the bit inverter 46 and then added by the adder 45 to the video signal 47 of the moving object. By such processing of the background removal circuit 9,
The difference between the video signal of the moving object and the background video signal is obtained, and an image of the moving object is created by removing the background video signal from the video signal of the moving object. The image signal of only the moving object created here is binarized by the binarization circuit 7 and then sent to the superposition processing circuit 8. The superposition processing circuit 8 calculates the logical sum of the output signal of the binarization circuit 7 and the locus image signal stored in the locus image memory 3 by an OR operation circuit 44.
The new trajectory image signal output from the OR operation circuit 44 is stored in the trajectory image memory 3. This new locus image signal is obtained by adding the locus image signal stored in the locus image memory 3 with the image signal of only the moving object after a predetermined time output from the binarization circuit 7. The processes in the background removal circuit 9, the binarization circuit 7, and the superposition processing circuit 8 are repeated for the above-described predetermined time to sequentially store the trajectory image signal of the moving object in the trajectory image memory.

Next, the removal of the background video signal from the start point video signal and the end point video signal of the moving object will be described with reference to FIG. The start point image signal or end point image signal stored in the start point image memory 5 and the end point image memory 6 and the background video signal stored in the background image memory 4 are input to the background removal circuit 8.
The background video signal is bit-inverted by the bit inverter 46, and then added to the start point or end point image signal by the adder 45. By this processing, the difference between the start point or end point image signal and the background image signal is obtained, and the image signal of only the moving object is created by removing the background image signal from the start point or end point image signal. Then, the moving object starting point image signal of only the moving object from which the background image signal is removed is the starting point image memory 5
Then, the moving object end point image signal of only the moving object is stored in the end point image memory 6. The image signals stored in the background image memory 4, the start point image memory 5, and the end point image memory 6 can be converted into a binarized output signal 48 by using the binarization circuit 7 as needed.

Further, this embodiment will be described again with reference to FIG. Trajectory image memory 3, background image memory 4, starting point image memory 5,
End-point image memory 6, binarization output 4 in binarization circuit
8. The video and image signals such as the video signal 47 in the speed measuring device can be output to the monitor (monitor television) 16 via the video output circuit 15. The generated arbitrary image signal is passed through the communication control circuit 13 to the image analysis device 10
Then, the image analysis device 10 performs image analysis to obtain the velocity of the moving object. Then, the shape information, the position information, the size, the movement amount, the speed of the moving object obtained by the image analysis,
Information such as vectors is displayed on the display device 40 and the printer 4.
1, the plotter 42 and the disk device 43 can be used to output or record numerical information or graphic information.

Next, a method of image analysis in the image analysis device 10 will be described. It is assumed that the TV camera 1 captures an image of the moving object 50 moving as shown by an arrow 51 in a place including the background 49 as shown in FIG. TV
When the video signal 47 is captured by the camera 1 and processed by the trajectory processing circuit 2 as described above, as an output example thereof, the binarized background image signal 5 of FIG.
2, the trajectory image signal 53 binarized by removing the background of FIG. 4 (c), the moving object start point image signal 54 binarized by removing the background of FIG. 4 (d), FIG. The moving object end point image signal 55 binarized by removing the background of e) is obtained. These image signals are input to the image analysis device 10 via the communication control circuit 13. A method for obtaining the velocity of a moving object using these image signals will be described.

First, the moving object start point image signal 54 and the moving object end point image signal 55 are analyzed. In an image including a plurality of moving objects, pixels are tracked along the contour constituent points of the image of each moving object, and the image signal of each moving object is distinguished. Next, the position of the center of gravity of the moving object start point image signal and the moving object end point image signal is calculated using the coordinate values of the contour constituent points, and the coordinates of each center of gravity are used as the starting point and end point coordinates of the moving object. An example of the method of calculating the center of gravity is as follows.

Here, (Xi, Yi) is the coordinate value of the N contour constituent points, and (Xg, Yg) is the coordinate of the center of gravity.

Next, the trajectory image signal 53 is analyzed. Pixels are tracked along the contour constituent points of the trajectory image signal 53 of each moving object,
The trajectory image signals of each moving object are recognized separately. Then, the position of the center of gravity is calculated using the above equations (1) and (2).
The area surrounded by the contour constituent points of the trajectory of each moving object is searched, and the barycentric coordinates of the start point and the end point of the moving object existing within the coordinate values of the contour constituent points are obtained. In the trajectory of the moving object, when the number of barycentric coordinates of the starting point and the ending point of the moving object are respectively plural, it is recognized that the trajectory is formed by a plurality of overlapping moving objects, and the starting point and the ending point of the moving object are recognized. Separate and recognize using coordinates. By the above process, the velocity at the moving average position of the moving object can be calculated as follows for the barycentric coordinate values of the start point and the end point of the moving object corresponding to the trajectory of the moving object.

Where u is the velocity component in the x direction, υ is the velocity component in the y direction, and X is the velocity component.
ge is the X-direction center-of-gravity coordinate value of the end point of the moving object, Xgs is the X-direction center-of-gravity coordinate value of the starting point of the moving object, Yge is the Y-direction center-of-gravity coordinate value of the ending point of the moving object, and Ygs is the starting point Y of the moving object.
The direction centroid coordinate value, Δt, is the time required for the moving object to reach the end point from the start point.

A normal TV camera has a finite scanning speed for one screen.
For example, the scanning speed is 1/60 second / screen for an industrial TV camera and 1/30 second / screen for a home TV camera. When the scanning time of one screen of the TV camera is T and the length of the moving object in the moving direction is L, when the moving object moves over the length L during the scanning time T of one screen, the moving object moves. The images of the moving object at the respective times at the predetermined time intervals, which are repeated in the process of creating the track of the object (process in the track creating circuit 2), do not overlap. The above-mentioned analysis method of the trajectory image is a method when the images of the moving object at each time continuously overlap, and when the images do not continuously overlap and become discontinuous, another analysis method is required. Become. The conditions under which the locus is continuous and discontinuous are as follows.

If V ≦ L / T, continuous trajectory. If V> L / T, discontinuous trajectory. Here, V is the magnitude of the velocity of the moving object. The image analysis device 10 determines whether the trajectory is a continuous trajectory or a discontinuous trajectory. Based on this determination result, the image analysis device 10 selects either the analysis method for the continuous trajectory described above or the analysis method for the discontinuous trajectory described below, and performs image analysis according to the selected analysis method. . FIG. 5 illustrates discontinuous trajectories of a plurality of moving objects. With a discontinuous trajectory,
The image signal 56 of the moving object at each time does not overlap the images of the moving object at the preceding and succeeding times, and becomes an image that coexists in one screen. Next, a method of analyzing such a discontinuous trajectory image signal will be described. First, pixels are tracked along the contour composing points of each discontinuous trajectory image signal, and the image signals of each moving object are distinguished. Those in which the center of gravity coordinates of the start point and the end point are included in the contours of these moving objects are stored as the start point and end point image signals of each moving object (corresponding memories 5 and 6 as described above). The barycentric coordinates are calculated and stored from the coordinate values of the contour constituent points of each moving object using equations (1) and (2). Next, using the barycentric coordinates of each moving object and the barycentric coordinates of the start point and the end point obtained in this way, the same moving object is identified from the image signal of the moving object included in one screen at each time. Will be described with reference to FIG. In Figure 6, three of the moving object A, the barycentric coordinates _A and B, and _{C 0 58, A 1 59,} A 2 6
0, A ₃ 61, B ₀ 62, B ₁ 63, B ₂ 64, B ₃ 6
5, and C ₀ 66, C ₁ 67, C ₂ 68, C ₃ 69 are present. Each moving object A, B and C has a barycentric coordinate A ₀
58 to barycentric coordinates A ₃ 61, barycentric coordinates B ₀ 62 to barycentric coordinates B ₃ 65, barycentric coordinates C ₀ 66 to barycentric coordinates C ₃
Assume that you have moved to 69. That is, the barycentric coordinate A ₀ 58,
B ₀ 62 and C ₀ 66 are the barycentric coordinates of the starting points of the moving objects A, B and C, and barycentric coordinates A ₃ 61, B ₃ 65 and C ₃ 69 are the barycentric points of the ending points of the moving objects A, B and C, respectively. Represents coordinates. In this case, the number of times the superposition processing circuit 8 has performed the superposition processing is four, and each moving object A,
Four B and C image signals are also included.
The procedure for identifying the image signal of the moving object A will be described below as an example.

(A) First, the barycentric coordinate B ₁ 63, which is the shortest distance from the barycentric coordinate A ₀ 58 of the starting point, is checked. Vector from barycentric coordinates B ₁ 63 A search is made in the vicinity of a point (small radius ε57) that is distant by a distance. In this example, since there is no corresponding point, the process moves to the search for the next point.

(B) The next closest barycentric coordinate B ₀ 62 is checked. Since this barycentric coordinate is the starting point of the moving object B, the search is stopped.

(C) determine the center of gravity coordinate _A 1 59. Vector from the center of gravity coordinate _A 1 59 Extended, barycentric coordinates A ₂ 60 is found when searching the neighborhood. Barycentric coordinates _A 1 59 barycentric coordinates _A 2 60, is regarded as the centroid coordinate _A 1 59 barycentric coordinates _A 0 58, from (i)
Perform the procedure up to (C). By repeating the same operation for the number of times of the superimposition processing, the barycentric coordinates A ₃ of the end point of the moving object A ₃
Can reach 61. For moving objects B and C, the image signal of the moving object at each time can be obtained by the same method. The Euler velocity is obtained from the distance between the start point and the end point of the moving object identified in this way, and the Lagrangian velocity can be obtained from the distance between the centers of gravity of the same moving object at each identified time. Further, if the small radius ε57 used in the search range in the above-described search method is a window with εx and εy on each side, the processing in the image analysis device and the computer becomes easier.

Next, the effect of this embodiment will be described. According to the present embodiment, it is possible to obtain a clear trajectory image signal that is not traced even if the background is removed when the trajectory image signal is created. Further, it is possible to perform the process of removing the background from the image of the measurement start point and the measurement end point including the image signals of the start point and the end point of the moving object, without providing a dedicated arithmetic mechanism for this process,
The calculation mechanism such as the background removal circuit 9 and the binarization circuit 7 used in the trajectory generation mechanism can be shared. This facilitates identification of the trajectory image signal of the moving object and the moving object start point and end point image signals, and improves the speed measurement accuracy and measurement efficiency. FIG. 7 is an example in which the velocities of a plurality of moving objects are measured by the moving object speed measuring apparatus of this embodiment.

By connecting the image storage medium provided in the trajectory creating device and the mechanism for creating the trajectory of the moving object in parallel, the computing mechanism of the mechanism for creating and printing the trajectory is shared, in view of computation between the image storage media. Since it can be used, it can be efficiently used without increasing the number of calculation mechanisms.

Furthermore, since the time-series image data necessary for obtaining the non-steady-moving velocity of the non-steady-moving object can be stored in a plurality of image recording media at high speed, an analyzer for obtaining the velocity of the moving object can be provided. Even when the velocity of the moving object cannot be obtained by the reactor im, it is possible to obtain the velocity of the object moving unsteadily that changes non-steadily.

Next, FIG. 12 shows another embodiment of the present invention. In this embodiment, the moving object velocity measuring device in the configuration of FIG.
An image memory 72 capable of storing a plurality of image signals by inputting the image signal 71 from the image output circuit 15 is provided. If a video signal is used as the image signal 71,
Image signals can be transferred to the image memory 72 at a high speed of 1/60 second, and high-speed acquisition of time-series image data necessary for obtaining the non-steadily changing speed of an object that moves non-steadily is possible. Become.

FIG. 13 shows an example of a moving object velocity measuring device in the case of measuring a three-dimensional velocity. In this embodiment, two TVs are used to capture the image of a moving object in stereo.
The camera 1 is provided with a video input circuit 14 and a video output circuit 1.
5, a binarization circuit 7, a superposition processing circuit 8, a background removal circuit 9,
The locus image memory 3, the background image memory 4, the starting point image memory 5 and the ending point image memory 6 each have two sets, each of which is a CPU 1
It is connected in parallel to the bus 12 connected to 1. By simultaneously processing the video signals output from the two synchronized TV cameras 1 in parallel, a locus image signal, a moving object start point image signal, and a moving object end point image signal captured stereoscopically from two directions are created. Then, the image is analyzed by the image analysis device 10. Thereby, the three-dimensional velocity of the moving object can be obtained.

Another embodiment of the present invention is a moving object velocity measuring device,
It is shown in FIG. In this embodiment, the function of the locus creating circuit 2 in the apparatus shown in FIG.
It was achieved by U11. The flow chart of FIG. 14 shows a processing procedure performed in the CPU 11. The hardware configuration of this embodiment is obtained by deleting the trajectory generation circuit 2 from the embodiment of FIG. First, the CPU 11
In order to set the necessary number of times of weight and the binarization level in the following process before executing the processing trouble of FIG. 14, the operator first sets the number of times of weight and the level of binarization to the CPU 11 as input data. input. Next, the video signal of the background photographing device which is the output of the TV camera 1 is input, and the video signal is digitized in the CPU 11 and then stored in the background image memory 4 as background image data. The video signal input in the case of the first weight processing is digitized as a start point image signal and stored in the trajectory image memory 3 as start point image data. The video signal input in the final superimposing process is digitized as the end point image signal and stored in the trajectory image memory 3 as the end point image data. TV
The video signal including the moving object and the background is sent from the camera 1 to the CPU.
When it is input to 11, it is processed as follows. This video signal, after being digitized, is determined not to include only the background video signal, and the background image memory 4
The background image data stored in is removed. That is, the digitized image data input from the TV camera 1 and the background image data are calculated, the static components contained in both image data are erased, and the background is removed to leave only the moving component. Then, the moving object trajectory image data obtained by performing the background removal in this way is converted into binary image data at a preset binarization level, and stored in the binarized image data and the trajectory image memory 3. It is logically ORed with the moving object locus image data being displayed. The moving object locus image data after this logical sum is stored in the locus image memory 3 again. Then, it is judged whether or not the superimposing process is completed, and if it is not completed, the process is returned to the previous end point input judgment part and the above-mentioned processes are repeated until the number of times of superimposition is completed. If it is finished, the image data of the background, the start point, the end point, the locus, etc. is transferred to the image analysis device, and the image is analyzed by the above-mentioned method to calculate the velocity of the moving object.

FIG. 15 shows another embodiment. In this embodiment, the color TV camera 73, the video input circuit 14, the superposition processing circuit 8 are provided.
And a background removing circuit 9, a locus creating mechanism 2, a locus color image memory 74, a background color image memory 75, a starting point color image memory 76, an ending point color image memory 77,
It is composed of a communication control circuit 13, a color image analysis device 78, and a monitor 16 for outputting the produced image as a video. The moving object velocity measuring apparatus of the embodiment shown in FIG. 1 is configured to be capable of processing a color image, and background removal and superimposition processing is performed on a video signal of a color moving object to obtain a color image. A trajectory image signal, a moving object start point image signal, a moving object end point image signal, etc. are created. According to the moving object velocity measuring apparatus capable of processing the color image of the present embodiment, the color of the moving object is recognized by the color image analyzing apparatus, and when there are a plurality of moving objects, it becomes easy to distinguish the individual moving objects. The identification of the start point, the end point, and the trajectory image becomes easier. Also, by using a temperature-sensitive liquid crystal object coated as a moving object or a temperature-sensitive liquid crystal object itself, by recognizing the color of the moving object, the temperature of the field in which the moving object exists or the temperature distribution can be determined. Alternatively, it is possible to measure simultaneously with the velocity distribution.

〔The invention's effect〕

According to the present invention, even when there are a plurality of loci of moving objects, the loci of the corresponding moving objects can be easily and accurately specified using the moving object starting point image data and the moving object end point image data. In particular, even when the trajectories of a plurality of moving objects intersect, it is possible to easily specify the corresponding trajectory. Therefore, since the moving speed of the corresponding moving object is obtained, the moving speed can be obtained efficiently.

[Brief description of drawings]

FIG. 1 is an embodiment of a block diagram of a moving object velocity measuring apparatus according to a preferred embodiment of the present invention, and FIG. 2 is an explanatory view showing a signal flow at the time of creating a locus image signal in FIG. FIG. 3 is an explanatory view showing a signal flow when a moving object start point image signal and a moving object end point image signal in FIG. 1 are created, and FIG. 4 (a) is an explanatory view of a moving object video signal including a background. Four
(B) to (e) are explanatory diagrams of the signal of the image output from the trajectory generation circuit, FIG. 5 is an explanatory diagram of trajectory images of a plurality of moving objects that are discontinuous trajectories, and FIG. 6 is a discontinuous trajectory. 7 is an explanatory view of the identification method in the case of FIG. 7, FIG. 7 is an explanatory view showing the measurement result of the velocity of the moving object measured by the apparatus of FIG. 1, and FIG. 8 is a background using the background image of the measurement starting point. FIG. 9 is an explanatory diagram of the trajectory image after removal, and FIG. 9 is an explanatory diagram of the trajectory image in which the background is removed by taking the difference between two consecutive frames.
FIG. 10 is an explanatory view of a starting point image in which the background is not removed,
FIG. 11 is an explanatory view of the end point image in which the background is not removed,
12 to 15 are configuration diagrams of another embodiment of the present invention. 1 ... TV camera, 2 ... trajectory creating circuit, 3 ... trajectory image memory, 4 ... background image memory, 5 ... starting point image memory, 6 ... endpoint image memory, 7 ... binarization circuit, 8 ......
Superimposition processing circuit, 9 ... Background removal circuit, 10 ... Image analysis device, 11 ... CPU, 12 ... Bus, 13 ... Communication control circuit, 14 ... Image input circuit, 15 ... Image output circuit, 16 ……monitor.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Katsuhiko Sagae 1168 Moriyama-cho, Hitachi-shi, Ibaraki Energy Research Laboratory, Hiritsu Seisakusho Co., Ltd. (56) References JP-A-54-44420 (JP, A) JP-A-58 -123462 (JP, A) JP-A-61-82170 (JP, A)

Claims (1)

[Claims] 1. A moving object which removes the background from a video signal including an image of a moving object and a background which is an output signal of a photographing device, and obtains the speed of the moving object based on the video signal from which the background is removed. In the speed measuring device, means for obtaining moving object starting point image data at a starting point of movement of the moving object during photographing by the photographing device, and moving object end point image data at end point of movement of the moving object during photographing. Means, a first memory for storing the moving object start point image data, a second memory for storing the moving object end point image data, and a trajectory of the moving object based on the video signal with the background removed. And a moving object for obtaining a velocity by using a locus creating means for obtaining the moving object, the moving object starting point image data, the moving object end point image data, and the moving object locus. And a means for determining the speed of the moving object based on the data of the specified trajectory.