CN201535753U

CN201535753U - Train rim crack detection system

Info

Publication number: CN201535753U
Application number: CN2009202395626U
Authority: CN
Inventors: 郭三华; 曹丽娟; 柳光日
Original assignee: Individual
Current assignee: Individual
Priority date: 2009-10-10
Filing date: 2009-10-10
Publication date: 2010-07-28
Anticipated expiration: 2019-10-10

Abstract

The utility model provides a train rim detection system, which comprises a UV light source, a CCD camera, a rim, an image collector, a background computer, output and alarm, wherein the CCD camera is connected with the image collector, the image collector is connected with the background computer, the background computer is used for outputting and alarming, the rim with the UV light source is irradiated by the CCD camera, and by utilizing an image processing technology in the image collector, the detection result is output and alarmed on the background computer. The system can detect whether the surface of the train wheel has cracks or not, has characteristics of real-time and high efficiency, can accurately detect whether the surface of wheel has cracks or not, avoids tedious artificial detection and human error, and increases work efficiency.

Description

Train wheel rim crack detection system

Technical field

The utility model provides a kind of train wheel rim crack detection system, belongs to the engineering inspection field.Specifically, the utility model is to gather the train wheel appearance (the magnesium powder has its effect of being corroded that prevents) that is coated with full magnesium powder on the production line in real time by camera, utilize image acquisition device, pass through image processing techniques, method such as adopt that the optimal threshold of Image Edge-Detection, morphological image computing, image is cut apart, train wheel is carried out feature detection, obtain final test result, and carry out Realtime Alerts.Have characteristics such as real-time, accuracy, avoided the loaded down with trivial details of manual detection, and the people is the error that causes.The accuracy height improves detection efficiency greatly.

Background technology

In the appearance wheel rim inspection side process of train wheel, traditional mode is exactly a manual detection, and error is many, and quantities is bigger, and efficiency ratio is lower.Wasted great amount of manpower, detecting quality can not guarantee.In the case, propose train wheel rim crack detection system, changed traditional detection mode, can save great amount of manpower, and detect quality and guaranteed, improved efficient, strengthened accuracy.

Summary of the invention

In view of above-mentioned, the utility model fundamental purpose is to provide the system between a kind of train wheel rim crack detection, satisfies real-time, the accuracy of train wheel rim crack detection on the production line, improves the efficient on the whole production line on the whole, is very practical.

In order to achieve the above object, the technical scheme that the utility model is taked is: the CCD camera is connected with image acquisition device, and image acquisition device is connected with background computer, utilizes background computer to export, report to the police.Shine above the wheel rim that adds ultraviolet source by the CCD camera, utilize the image processing techniques in the image acquisition device, testing result is exported on background computer, reported to the police.When utilizing image processing techniques to handle, comprise edge of image detection technique, binaryzation technology and optimal threshold and technology such as cut apart for the image of real-time collection.Concrete grammar is: utilize the Canny operator to detect to collect the image border, thereby can extract continuous and complete edge.And then, use level and smooth, the refinement thick edge of burn into thinning algorithm in the morphology again with morphologic binaryzation, the gap between the elimination dual edge of expanding, make edge more approaching former target on width to obtain the first pending width of cloth image; Original image is carried out Fourier transform handle, cut apart with optimal threshold again and obtain second result image, then two width of cloth images are carried out point multiplication operation, use template to remove residual noise at last, obtain final crack detection result.With real-time the showing on background computer of image after whole monitoring image and the processing, if having the train wheel of crackle, system has improved work efficiency greatly with Realtime Alerts.

Beneficial effect

The utility model utilizes the gordian technique in the Flame Image Process, and the video monitoring image on the train wheel that collects is in real time handled, and obtains desirable detection effect, increases work efficiency, and the accuracy height has very strong actual effect.

Description of drawings

Fig. 1 is the utility model system principle diagram.Described system mainly comprises ultraviolet source (1), CCD camera (2), wheel rim (3), image acquisition device (4), background computer (5), report to the police (6).

Fig. 2 is the real-time testing process figure of the utility model.Mainly comprise ultraviolet source (1), CCD camera (2), wheel rim (3), image acquisition device (4), background computer (5), up-to-standard (7), (8) off quality totally 6 steps.

Fig. 3 is an image processing techniques process flow diagram in the image acquisition device in the present utility model system.Mainly comprise the original image (9) that obtains, burn into thinning algorithm (12) in binaryzation, expansion algorithm (11), the morphology in Canny rim detection (10), the morphology, first is handled, and image (13), Fourier transform (14), optimal threshold are cut apart (15), second result figure (16), point multiplication operation (17), template denoising (18), the image (19) handled totally 11 steps.

Embodiment

Below in conjunction with accompanying drawing the present utility model is further detailed.

Fig. 2 is the real-time testing process figure of the present utility model, and as shown in Figure 2, its concrete steps are as follows:

Step 1 ultraviolet source (1) shines on the train wheel rim that is coated with full magnesium powder, and the purpose of utilizing ultraviolet source irradiation is that the realtime graphic that camera is obtained has very strong comparative;

Many CCD cameras of step 2 (2) are monitored in real time to wheel rim (3), are convenient to follow-up real-time processing like this; Will guarantee in this process that wheel rim (3) rotates on production line, a plurality of CCD cameras (2) will be controlled the real-time collection of different piece image of the wheel rim (3) of rotation respectively, will be with different CCD cameras as the aspects such as the outside of the side of wheel rim, wheel rim;

Step 3 utilizes image acquisition device (4) to carry out the real-time processing of image, utilizes technique of image edge detection, binaryzation technology, optimal threshold cutting techniques that real-time images acquired is handled;

Realtime graphic that step 4 is obtained and real-time testing result image are delivered to background computer (5) display result;

If step 5 is not found crackle, then think rim surface up-to-standard (6), if find crackle, then to report to the police, think that like this rim surface quality has problem, (7) off quality.

Fig. 3 is an image processing techniques process flow diagram in the image acquisition device in the present utility model system, and as shown in Figure 3, its concrete steps are as follows:

Step 1 utilizes Canny rim detection (10) algorithm to detect the edge of the original image (9) that obtains, thereby can extract continuous and complete edge;

Step 2 utilizes morphologic binaryzation, expansion algorithm (11) to eliminate the gap between the dual edge

Step 3 is used burn into thinning algorithm smoothing algorithm (12) again, the refinement thick edge, make the edge on width more near original object, obtain first result figure (13);

The step 4 pair original image that obtains (9) carries out Fourier transform and handles (14), cuts apart (15) with optimal threshold again and obtains second result image (16);

Step 5 is used template denoising (18) method, the image that obtains handling (19) at last then to two width of cloth picture point multiplications (17).

Claims

1. train wheel rim crack detection system, its design feature is: described system is made up of ultraviolet source (1), CCD camera (2), wheel rim (3), image acquisition device (4), background computer (5), output, report to the police (6), CCD camera (2) is connected with image acquisition device (4), image acquisition device (4) is connected (5) with background computer, background computer (5) export, report to the police (6).

2. according to claims 1 described a kind of train wheel rim detection system, it is characterized in that: shine by CCD camera (2) above the wheel rim (3) that adds ultraviolet source (1), utilize the image processing techniques in the image acquisition device (4), testing result is gone up output, warning (6) at background computer (5).