JPH05126686A - Vehicle inspecting device - Google Patents

Abstract

PURPOSE:To photograph an image over the whole periphery of a vehicle and automatically perform the process corresponding to the visual inspection of a human with a computer by employing a special photographing system using a one-dimensional sensor camera from the front against the incoming vehicle. CONSTITUTION:When a wheel to be inspected is advanced while being rotated toward a one-dimensional sensor camera installed horizontally so that the center section of an incoming vehicle enters the center of the visual field, image signals corresponding to horizontal lines 1, 2,...n-1, (n) on the surface of the wheel apart by the rotation angle da radian of the wheel at each signal extraction interval dt of the camera 1 are sent to an image processor 5 from the camera 1 in sequence. When the image signal data corresponding to the horizontal lines 1, 2,...n-1, (n) are two-dimensionally arranged in the image processor 5, the deployed image data of the wheel surface for one circumference are obtained. The deployed image data of the wheel surface are analyzed by the image processor 5, and abnormalities such as flaws and objects existing on the surface of the wheel can be discovered.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle, such as a railway, whose traveling route of wheels is fixed along a track.
This is for non-contact inspection of scratches and inclusions on the surface of the wheel passing by using an inspection device on the ground. The present invention can be applied to general automobiles by restricting the route when passing in front of the inspection device. It can also be applied to the tire inspection during low-speed wheel running of the levitation railway, which is currently under development.

[0002]

2. Description of the Related Art In railways, scratches on the wheel surface caused by sliding due to sudden braking are called flats, which are a major cause of vibration and noise during running. It is necessary to deal with it by cutting back.

As the inspection method therefor, the method of visually inspecting the wheels of the vehicle stopped at the driving base by a human is the most reliable method, but it is very time-consuming and
Only a part of the entire circumference of the wheel can be confirmed in the stopped state. On the other hand, as a simple method, a method has been used in which an inspector on the ground or on a vehicle perceives vibrations and noise generated when a wheel with scratches travels, and evaluates their presence and degree. It was

In contrast to such a method relying entirely on human sensation, recently, as an automatic inspection means, the vibration generated when a damaged wheel is running is detected by a vibration sensor attached to a rail or sleeper, and A method of evaluating the size is partially used.

[0005]

As the evaluation of the scratches on the wheel surface, the most reliable one is to visually check the stopped wheel by a human.
The size and depth are evaluated, but as mentioned in the previous section, it is very time-consuming and only a part of the wheels can be confirmed in the stopped state. In the latter method using the vibration sensor, the degree of vibration changes depending on the speed of the traveling train, the shape of the scratch, the elapsed time from the scratch, etc., and it is difficult to correspond to the actual size of the scratch. There is a problem that the reproducibility of vibration evaluation is also low. Therefore, it is an object of the present invention to solve the problem by capturing an image of the entire circumference of a wheel with a special image capturing device and automatically performing a process corresponding to a human visual inspection using image processing by a computer. It is a task to do.

[0006]

In order to solve the above-mentioned problems, first of all, by taking a special photographing method using a one-dimensional sensor camera from the front of the approaching wheels,
It is possible to obtain a developed image of one round of the wheel surface. Further, by using an image processing computer for this developed image, it becomes possible to automatically perform an all-around inspection of all passing wheels without requiring manpower.

[0007]

At the center of the means for solving the problems in the present invention is a special photographing system using a one-dimensional sensor camera. The basic operation of the shooting method is that a part of the wheel surface moving while rotating is continuously shot from the front by a one-dimensional sensor camera, and the video signal data is arranged in time series. It is based on the function of the principle of operation that image data obtained by expanding the surface over the entire circumference can be obtained. By this basic action, a moving wheel is not stopped and a wheel surface image is obtained over the entire circumference. By processing the image data by a computer, it is possible to automatically determine the scratches on the wheel surface. I can.

[0008]

An embodiment of the present invention will be described with reference to the drawings. FIG. 1 shows an example of the overall configuration when the present invention is applied to a railway wheel inspection. That is, a certain section of the track on which the vehicle is traveling is set as a wheel surface inspection section, and a lighting device 2 for illuminating the wheel surface is installed in that section, and a wheel 3 to be inspected of a vehicle entering this section. Illuminate. A pair of one-dimensional sensor cameras 1 are installed near the front rails of the advancing wheels, and horizontal image signals from the one-dimensional sensor cameras 1 are continuously supplied to the image processing device 5 while the wheels pass through this section. By capturing and accumulating this in the image accumulating device 6, a developed image over the entire circumference of the wheel on the wheel surface is obtained. The developed image data is subjected to image processing by the image processing device 5, and it is automatically determined whether or not there are scratches or inclusions on the wheel surface which are a problem in wheel management.

FIG. 2 and FIG. 3 are explanatory views of the action of obtaining a developed image when a moving wheel is monitored by a one-dimensional sensor camera fixed on the ground. When the wheel to be inspected rotates while advancing with respect to the horizontally installed one-dimensional sensor camera 1 such that the center of the advancing wheel enters the center of the field of view as shown in FIG. 2, In the visual field area of the one-dimensional sensor camera 1, horizontal lines 1, 2, 3 ,. ． ． The video signal corresponding to n-1, n is
Sequentially sent from the camera to the computer. This horizontal line 1,
2, 3 ,. ． ． When the video signal data corresponding to n-1 and n-1 are arranged two-dimensionally in a computer, two-dimensional developed image data of the wheel surface as shown in FIG. 3 is obtained. At this time, the traveling speed v (m / s) of the wheel, the signal acquisition interval dt (s) of the camera, and the number of horizontal lines per wheel n
The relationship with is as follows, where r (m) is the radius of the wheel.

[0010]

[Equation 1]

This is an explanation of the case where the camera is installed in front of the wheel. However, in reality, when the camera is installed in front of the wheel, the one-dimensional sensor camera may come into contact with the wheel as it advances. Then, when inspecting the surfaces of a plurality of wheels, there is a problem that the second and subsequent wheels are hidden by the first wheel. Therefore, as shown in FIG. 1, the one-dimensional sensor camera is moved slightly outward from the front of the track. Install it in the appropriate position.

When the one-dimensional sensor camera is installed slightly offset from the front of the wheel, the image data shown in FIG. 4 can be obtained by storing the image data in the computer by the above method. This can be processed in an image processing computer and converted into a front developed image corresponding to FIG.

An example of a processing procedure therefor will be described with reference to FIG. First, in FIG. 5A, a quadrilateral portion A and a triangle portion B surrounded by a vertical line from a point n lines ahead of the number of lines for one round from a line where a complete video signal is obtained. To obtain a parallelogram by connecting them to each other as shown in (b), and further shearing the parallelogram in a horizontal direction to form a right-angled rectangle as shown in (c). An image of the developed surface of the wheel is obtained.

Further, FIG. 6 shows the conditions under which the entire circumference of the wheel to be inspected in the wheel inspection system according to the present invention falls within the viewing angle α of the one-dimensional sensor camera 10 and a clear image is obtained over the entire circumference. 9 shows the conditions for the depth of field 9 of FIG. That is, the viewing angle α is required that the viewing angle α of the one-dimensional sensor camera covers the movement of the wheel in one rotation in the inspection section indicated by the movement from point a to point b in FIG. Therefore, the condition that the lens system of the one-dimensional sensor camera covers the same range is a condition for the lens system to obtain a clear image of the entire circumference.

Further, FIG. 7 shows a case where there are a plurality of wheels such as a bogie of a railway vehicle, because these can be separated in the primary image data from the one-dimensional sensor camera as shown in FIG.
The conditions of the camera installation position are shown. In other words, the angle of attack β of 11 formed by the optical axis 13 of the camera with respect to the rail on which the wheel enters must be larger than the condition of FIG. 7.

[0016]

When the present invention is applied to the railway wheel inspection, the wheel inspection work, which has required a lot of manpower in the conventional work system under the unfavorable working environment, is completely automatic. It becomes possible to perform work, and labor saving of large work and improvement of working environment are realized.

Furthermore, by using a computer-based inspection method based on a unified inspection standard, conventional manual inspection standard variations are eliminated, and the inspection results are stored in a database to manage time series history. Will also be possible. In addition, the reliability of the inspection is improved by implementing all-around inspection of all wheels, and by utilizing these comprehensively, the facility maintenance function is significantly improved and the ride comfort for vehicle users is improved. You can

Further, by applying the present invention to a wheel inspection of an automobile, for example, detection of abnormal wheels at the entrance of a highway, automatic inspection at the time of entering / leaving a business office of a public bus, and the like, road traffic safety can be improved. It is expected to contribute to improvement.

Further, by applying it to a system for detecting abnormalities of rubber tires for low speed running of a magnetic levitation railway currently under development, it contributes to improving the safety of the magnetic levitation railway and speeding up its practical application. There is expected.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram when the present invention is applied to a railway wheel inspection.

FIG. 2 is a diagram illustrating scanning of a wheel surface.

FIG. 3 is a diagram showing continuous images when images are taken from the front.

FIG. 4 is a diagram showing continuous images taken obliquely from the front.

FIG. 5 is a diagram illustrating conversion from an oblique front image to a front image.

FIG. 6 is a diagram illustrating a condition of depth of field.

FIG. 7 is a diagram illustrating a condition of an attack angle.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 1-dimensional sensor camera 2 Illumination device 3 Wheel to be inspected 4 Orbit 5 Image processing device 6 Image storage device 7 1-dimensional sensor 8 Moving distance for one rotation 9 Depth of field 10 Viewing angle α 11 Angle of attack β 12 β-α / 2 13 Optical axis of camera 14 Adjacent wheel

Claims (1)

[Claims] 1. A one-dimensional sensor camera is installed in the vicinity of the track in order to detect abnormalities on the wheel surface of a vehicle traveling on the track, and the one-dimensional sensor camera continuously photographs the advancing wheel surface. Then, the wheel surface development image data for one round is acquired, the wheel surface development image data is analyzed by an image processing device, and abnormalities such as scratches and inclusions existing on the wheel surface are found. apparatus.