Specific embodiment
The feature and exemplary embodiment of the various aspects of the utility model is described more fully below, in order to keep this practical new
The objects, technical solutions and advantages of type are more clearly understood, and with reference to the accompanying drawings and embodiments, carry out the utility model into one
Step detailed description.It should be understood that specific embodiment described herein is only used for explaining the utility model, it is not used to limit this reality
With novel.To those skilled in the art, the utility model can not need some details in these details
In the case where implement.Below the description of embodiment is used for the purpose of providing by showing the example of the utility model to this reality
It is better understood with novel.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that the process, method, article or equipment for including a series of elements not only includes those elements, and
And further include other elements that are not explicitly listed, or further include for this process, method, article or equipment institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence " including ... ", it is not excluded that including described want
There is also other identical elements in the process, method, article or equipment of element.
A kind of tunnel defect feature detection system is provided in the present embodiment, and Fig. 1 is according to the utility model embodiment
Tunnel defect feature detection system structural schematic diagram, as shown in Figure 1, the system include: battery 1, mobile platform 2, control
Device 3 and image collecting device 4 processed, control device 3 and image collecting device 4 are mounted on mobile platform 2;Control device 3 divides
It is not electrically connected with mobile platform 2 and image collecting device 4, to control mobile platform 2 and image collecting device 4;Wherein, image
Acquisition device 4 includes: up-right support 41, camera fixing support 42 and strip source 43, wherein
Up-right support 41 is mounted on the center of mobile platform 2;
The main body of camera fixing support 42 is fan-shaped or disc, and the center of circle is surrounded in the main body of camera fixing support 42
Multiple mounting portions 421 are distributed in a manner of fan-shaped array, consolidate on some or all mounting portions 421 in mounting portion 421
Surely there is line-scan digital camera 5;Camera fixing support 41 is identical with the direction of motion of mobile platform 2 with the direction for passing through the axis in the center of circle
Form is mounted on the top of up-right support 41;
Strip source 43 is mounted on the top of up-right support 41 around the center of circle in a manner of fan-shaped array.
Using above system, multiple line-scan digital cameras are mounted on above-mentioned camera fixing support, it can be in one-time detection
The middle Image Acquisition for completing full section of tunnel, so that the tunnel defect feature detection system solved in the related technology can not be simultaneously
The problem of realizing the detection of full section of tunnel, improves tunnel defect feature detection efficiency.
Optionally, the capacity of battery 1 is determined according to the power of other electrical components and cruise duration, at least to meet
Electricity needed for Tunnel testing of tunnel defect feature detection system.
Optionally, mounting portion 421 includes camera stabilization holder, and line-scan digital camera 5 is fixed on camera by camera stabilization holder
On fixed bracket 42.Since sensibility of the line-scan digital camera for vibration is quite high, to guarantee to obtain high quality graphic, camera is consolidated
Due to camera stabilization holder, can effectively reduce vibration influences, and obtains high quality graphic.
Optionally, the field range for the whole line-scan digital cameras 5 being fixed on camera fixing support 42 can cover tunnel
Tunneling boring.Wherein, the tunneling boring in tunnel does not include the railway roadbed and track in tunnel.
Optionally, at least there is weight in the field range for being fixed on the two adjacent linear array camera 5 on camera fixing support 42
Folded region, to carry out the high-precision joining of image.Optionally, overlapping region is not less than 20cm.
The line-scan digital camera 5 and its accessory (such as camera lens etc.) installed on camera fixing support 42 can be according to obtaining
The difference (such as the radius in tunnel) of the required precision, measurand visual field size that take image is selected, and determines camera
Quantity;The suitable camera of resolution ratio and the number of camera are conducive to guarantee to obtain the precision of images item while save the cost
High quality graphic is obtained under part.All line-scan digital cameras 5 can share the same image pick-up card;In order to improve the reality of Image Acquisition
Shi Xing, it is preferable that each line-scan digital camera 5 uses an individual image pick-up card.
Optionally, line-scan digital camera 5 is high resolution industrial line-scan digital camera.High resolution industrial line-scan digital camera can be mentioned substantially
The resolution ratio of hi-vision, to improve detection accuracy.Also, the picture-taken frequency highest of industrial line-scan digital camera can achieve
66000Hz can guarantee the high efficiency for the image that industrial line-scan digital camera acquires during the motion.
Optionally, strip source 43 is highlighted bar-shaped LED light source, which is installed on camera with fan-shaped array manner and fixes
On bracket 42, for guaranteeing that light is uniform and is covered with whole field ranges of more line-scan digital cameras 5.
Optionally, control device 3 further includes odometer 6.Odometer 6 is preferably high precision odometer, for guaranteeing linear array
High-precision corresponding relationship between the image and practical mileage that camera obtains.
Optionally, high-precision corresponding relationship between the image and practical mileage in order to guarantee line-scan digital camera acquisition, is being adopted
It is every by default fixed Image Acquisition distance, odometer 6 during carrying out Tunnel testing with tunnel defect feature detection system
It can choose and once proofreaded.
Optionally, whole line-scan digital cameras 5 that control device 3 is also used to be fixed on camera fixing support 42 are collected
Image is spliced, and the practical mileage measured according to odometer 6 carries out practical mileage and tunneltron to the image spliced
The mark of piece number.Spliced image is marked according to practical mileage and tunnel duct piece number, is convenient for Disease Characters positioning.Splicing
The image obtained afterwards just can be used in manual or automated Disease Characters mark, analysis and statistics;It is based especially on industrial line
The high resolution features of array camera, the image obtained after splicing can mark, analyze and count the diseases such as the fine cracks in tunnel
Feature.
Optionally, control device 3 is for first acquiring whole line-scan digital cameras 5 according to the positional relationship between line-scan digital camera 5
Image tentatively spliced, then according to RANSAC algorithm to image carry out high-precision joining.Wherein, between line-scan digital camera 5
Positional relationship be to have been demarcated between line-scan digital camera 5 in advance according to determined by the calibration in advance to more line-scan digital cameras
Positional relationship, and line-scan digital camera 5 is installed on camera fixing support 42 according to identical positional relationship.
Optionally, control device 3 is also used to generate the triggering letter of line-scan digital camera 5 according to the practical mileage that odometer 6 measures
Number, to adjust the picture-taken frequency of line-scan digital camera 5 in real time according to the movement velocity of mobile platform 2.Wherein, all linear array phases
Control of the machine 5 all by the same trigger signal.By the above-mentioned means, tunnel defect feature detection system can be in movement velocity
The image of high quality is acquired in situation heterogeneous.For example, tunnel defect feature detection system can be set to mobile platform 2 often
It moves 0.3mm and acquires an image, then when the speed that odometer 6 measures mobile platform 2 is fast, the touching of the generation of control device 3
The frequency of signalling is just high, when the speed that odometer 6 measures mobile platform 2 is slow, the first draft signal of the generation of control device 3
Frequency accordingly reduces, to realize that every 0.3mm acquires an image.
Optionally, the mobile platform 2 in the present embodiment has sufficiently rigid, and front and back has enough spans, so that moving
The stability of moving platform 2 during the motion, reducing vibration influences picture quality.
Optionally, control device 3 further includes display screen, is used for 6 acquired image of real-time display line-scan digital camera, and/or,
The image obtained after the splicing of control device 3.By display screen, can be observed in real time the working condition of line-scan digital camera 5 with
And its whether the image obtained meets quality requirement.
It should be noted that above-mentioned control device 3 also wraps other than can have the components such as display screen, odometer 6
Other basic element of character needed for realizing its function are included, such as: these basic element of character can be the industrial personal computer for calculation process, use
In the large capacity solid state hard disk etc. of storing data.
A kind of tunnel defect characteristic detection method is additionally provided in the present embodiment, and this method is applied to above-mentioned tunnel disease
Evil feature detection system.Fig. 2 is according to the flow chart of the tunnel defect characteristic detection method of the utility model embodiment, such as Fig. 2
Shown, which includes the following steps:
Acquisition picture size is preset in step S201, the initial position of uncalibrated image acquisition;
Step S202 starts tunnel defect feature detection system, and mobile platform is since initial position in the track in tunnel
On move forwards;
Step S203, while mobile platform moves forwards in orbit, control device sends trigger signal and gives linear array phase
Machine;
Step S204, line-scan digital camera acquire image according to trigger signal;
Step S205, control device carry out the whole line-scan digital camera acquired images being fixed on camera fixing support
Splicing;
Step S206, the practical mileage that control device is measured according to odometer carry out practical mileage to the image spliced
With the mark of tunnel duct piece number.
Optionally, it includes: control device according to being mounted on mobile platform that control device, which sends trigger signal to line-scan digital camera,
On odometer measurement practical mileage, to line-scan digital camera send trigger signal, with real-time according to the movement velocity of mobile platform
Adjust the picture-taken frequency of line-scan digital camera.
Optionally, control device spells the whole line-scan digital camera acquired images being fixed on camera fixing support
Connecing includes: first tentatively to be spliced according to the positional relationship between line-scan digital camera to the image that whole line-scan digital cameras acquire, then
High-precision joining is carried out to image according to RANSAC algorithm.
In order to keep the description of the utility model embodiment clearer, it is described and says below with reference to preferred embodiment
It is bright.
This preferred embodiment provides a kind of tunnel defect characteristic detection method, and Fig. 3 is preferably real according to the utility model
The flow chart of the tunnel defect characteristic detection method of example is applied, Fig. 4 is special according to the tunnel defect of the preferred embodiment in the utility model
Levy signal, the energy flow schematic diagram of detection method.With reference to Fig. 3 and Fig. 4, this method comprises the following steps:
Step S301 carries out camera model selection according to default detection accuracy, field range and camera number adjusts.
In the present embodiment, image resolution ratio precision by taking 0.3mm as an example (but being not limited to 0.3mm), tunnel diameter is with 5.5m
For (but it is not limited to 5.5m), the range that tunnel is detected isTherefore the visual field of tested range is
12.9525m needs 12.95 × 1000/0.3=43175 pixel in the case where guaranteeing that measurement accuracy can reach measurement request,
Therefore 4 resolution ratio can be used to be detected for 12288 × 1 industrial line-scan digital camera, single camera field range can achieve
3.6864m, the overlapping range between adjacent cameras can reach 0.6m.
Step S302, camera, light, odometer, industrial personal computer, battery etc. are fixed on carrier.
In the present embodiment, camera is installed on bracket with fan-shaped array manner, highlights bar-shaped LED lamp also with fan-shaped array side
Formula is installed on bracket, then secures a bracket to mobile platform, and odometer, industrial personal computer, display screen, battery are all fixed on shifting
Moving platform guarantees that enough rigidity prevent relative motion between each device, and LED lamplight is uniformly covered with the holistic vision model of camera
It encloses, as shown in Figure 5.
More industrial line-scan digital cameras are fixed on stabilization holder, and three axis of stabilization holder are adjusted by step S303,
Guarantee that camera center of gravity is overlapped with holder center.
Step S304, camera calibration.
In the present embodiment, in advance using Zhang Zhengyou calibration algorithm or other traditional calibration algorithms to phase before entering tunnel
Machine is demarcated, and the inside and outside parameter of camera is obtained, and determines the positional relationship between camera, is guaranteed between the adjacent cameras visual field extremely
The overlapping region of few 20cm, guarantees the required precision of image mosaic.
Figure position is adopted in step S305, in advance calibration starting.
When mobile platform is moved to starting detection position, starting mileage number is demarcated in advance and determines mileage number change direction
(become larger or reduce) guarantees to realize high-precision corresponding relationship between the practical mileage of image information and tunnel obtained, and in image
It is every primary by default fixed range odometer check and correction in collection process.
Step S306, pre-set image size adopt figure frame number and reach default size and save an image.
Parameter preset is adopted i.e. one image of preservation when figure frame number often reaches 10000 and (for 10000 but is not limited to
10000), and image is named with the practical mileage number in the corresponding tunnel of the whole story.
Step S307, odometer provide trigger signal.
In this example, the movement speed that mobile platform is arranged is 20km/h, and controller can be believed according to the real-time mileage of odometer
Breath generates trigger signal, and triggering camera is taken pictures, and adopts figure frequency size, the non-homogeneous feelings of mobile platform movement velocity without default
It still is able to obtain high quality graphic under condition;In image acquisition procedures, acquisition image can be observed in real time in display screen
Quality and camera working condition can find simultaneously to adjust in time in time as gone wrong in image acquisition process.
Step S308, according to the real-time mileage information of odometer, camera adjust automatically line-scan digital camera adopts figure frequency.
In the present embodiment, the acquisition precision of images is 0.3mm, odometer real-time transmission mileage information to industrial personal computer, industrial personal computer
The every mobile 0.3mm of camera, which is controlled, according to parameter preset triggers an Image Acquisition thread.The line-scan digital camera highest adopts figure frequency can
Up to 66000Hz, when obtaining the precision of images is 0.3mm, mobile platform highest movement speed is estimated up to 71km/h, can
See, under such high speed, image acquiring device can still obtain high quality graphic, and real-time transmission to image memory device is protected
It deposits.
Step S309, odometer check and correction.
In the present embodiment, every by pre- during carrying out Tunnel testing using tunnel defect feature detection system
If fixed Image Acquisition distance, odometer are once proofreaded.
Step S310, image mosaic.
It is after adopting figure, the image of each camera acquisition acquisition is preliminary according to the positional relationship progress image demarcated in advance
Splicing.Then two picture laps are detected according to RANSAC algorithm, and realized between adjacent two cameras acquisition image
At a high speed, high-precision joining.
Step S311 marks corresponding tunnel mileage or section of jurisdiction number on the image.Image after the completion of splicing is realized high-precision
Position mark is spent, the high-precision corresponding relationship between image and the practical mileage in tunnel or section of jurisdiction number is formed.
Step S312 carries out all kinds of Disease Characters marks, analysis and statistics on splicing complete image, can especially mark,
The small Disease Characters such as analysis and the similar fine cracks of statistics, directly mark out all Disease Characters, and analyze disease in original image
The size of the diseases such as chip off-falling, infiltration, crack in evil feature, such as fracture length and width are mentioned to tunnel safety maintenance
For reliable reference value.
In conclusion solving tunnel defect feature detection system in the prior art by the utility model embodiment
It is low with the detection efficiency of method, precision is low, mobile platform serious vibration influence picture quality the problems such as.
The above descriptions are merely preferred embodiments of the present invention, is not intended to limit the utility model, for this
For the technical staff in field, various modifications and changes may be made to the present invention.It is all in the spirit and principles of the utility model
Within, any modification, equivalent replacement, improvement and so on should be included within the scope of protection of this utility model.