阅读说明：本技术 一种基于摄影测量的无人机隧道变形检测方法 (It is a kind of based on photogrammetric unmanned plane tunnel deformation detection method ) 是由 胡启军 谭杰 何乐平 于 2019-09-30 设计创作，主要内容包括：本发明涉及无人机自动检测领域的一种基于摄影测量的无人机隧道变形检测方法,由数据采集系统、数据处理及分析系统和结果展示系统组成,其包括如下具体步骤：1)根据隧道检测区域设定无人机飞行的GPS起止坐标；2)无人机按照步骤1)设定的航线自动巡航,并通过搭载在无人机的激光发射装置发出垂直于隧道轴线的均匀明亮的激光,同时通过搭载在无人机上的图像获取装置对隧道激光轮廓进行快速记录；3)将采集的隧道激光轮廓进行拟合,结合空间定位装置,将拟合断面与前一次检测结果或者设计轮廓进行对比,计算出变形情况。本发明可通过无人机实现隧道变形的自动化检测,对隧道变形进行安全评估,并对后续可能的变化做出相应的判断。(The present invention relates to a kind of based on photogrammetric unmanned plane tunnel deformation detection method of the automatic detection field of unmanned plane, it is made of data collection system, data processing and inversion system and result display systems comprising following specific steps: 1) setting the GPS start-stop coordinate of unmanned plane during flying according to Tunnel testing region；2) the course line automatic cruising that unmanned plane is set according to step 1), and the laser beam emitting device by being mounted in unmanned plane issues the laser of the homogeneous and bright perpendicular to tunnel axis, while the image acquiring device by being mounted on unmanned plane carries out fast recording to tunnel laser profile；3) the tunnel laser profile of acquisition is fitted, in conjunction with space positioning apparatus, fitting section is compared with a preceding testing result or design profile, calculates deformation.The present invention can realize the automatic detection of tunnel deformation by unmanned plane, carry out security evaluation to tunnel deformation, and make corresponding judgement to subsequent possible variation.)

1. a kind of based on photogrammetric unmanned plane tunnel deformation detection method, mainly by data collection system, data processing and Analysis system and result display systems composition, which is characterized in that the UAV system detecting step are as follows:

1) the GPS start-stop coordinate of unmanned plane during flying is set according to Tunnel testing region；

2) unmanned plane enables vision system in the place that GPS signal is faint according to the course line automatic cruising that step 1) is set automatically It assists cruising with laser radar, and the laser beam emitting device by being mounted in unmanned plane is issued perpendicular to the uniform bright of tunnel axis Bright laser；

3) while unmanned plane issues laser according to step 2), by the image acquiring device that is mounted on unmanned plane to tunnel Laser profile carries out fast recording；

4) when the data acquisition of unmanned plane completion step 3), the tunnel laser profile of acquisition is fitted；

5) when the data processing for completing step 4), in conjunction with space positioning apparatus, will fitting section and a preceding testing result or Design profile compares, and shows deformation.

2. according to claim 1 based on photogrammetric unmanned plane tunnel deformation detection method, it is characterised in that: described GPS start-stop coordinate, including setting initial coordinate P₁, end coordinate P₂, the height of unmanned plane, speed, posture, course, guide nothing It is man-machine to fly by regulation axis.

3. according to claim 1 based on photogrammetric unmanned plane tunnel deformation detection method, it is characterised in that: described Laser beam emitting device, issue perpendicular to tunnel axis homogeneous and bright laser, consistently formed during unmanned plane during flying bright Aobvious tunnel contour lasing area.

4. according to claim 1 based on photogrammetric unmanned plane tunnel deformation detection method, it is characterised in that: described Image acquiring device, including four CCD cameras, and adjust angle between camera obtain best image Duplication, in unmanned plane In flight course, continuously taken pictures by setpoint frequency to laser section.

5. according to claim 1 based on photogrammetric unmanned plane tunnel deformation detection method, it is characterised in that: described Cross section correct algorithm be pass through least square method fitted ellipse section.

6. according to claim 1 based on photogrammetric unmanned plane tunnel deformation detection method, it is characterised in that: described Spatial position of the space positioning apparatus record unmanned plane in tunnel, the image location information that can be obtained with image acquiring device It is corresponding, it is quickly found out the specific location of tunnel deformation.

Technical field

The present invention relates to a kind of unmanned plane tunnel deformation detection methods, especially disclose a kind of based on photogrammetric nothing Man-machine tunnel deformation detection method belongs to unmanned plane automatic measurement technique field.

Background technique

Currently, China be completed and the tunnel that goes into operation no matter from quantity or total kilometrage for be reached compared with On a large scale, the maintenance stage has wherein greatly been entered.With continuing to increase for China tunnel quantity, tunnel structure The detection of operation state becomes particularly important.Tunnel deformation detection is work most time-consuming in current check maintenance work, highway Tunnel testing under normal circumstances can close traffic, influence running speed；Subway tunnel can only be in the finite time after stopping operation It is inside checked, the time is hasty, is easy error.It improves tunnel operation detection speed and has become necessarily becoming for Tunnel testing development Gesture.

For tunnel deformation detection, there are many different Method and kit fors.Total station is a kind of biography being widely adopted System technology, precision is high, but requires a great deal of time and carry out live adjustment and manually control.In recent years, 3D laser scanning skill Art is flourished in underground engineering, is capable of the threedimensional model in automatically scanning tunnel, high degree of automation, but scanning speed Slowly, at high cost.Nowadays, traditional close-shot photography measure technique is because its is at low cost, and precision is high and quickly grows, but index point It must install in advance, previous work is very time-consuming.Recently, the appearance of vehicular laser scanner improves tunnel examination and maintenance Efficiency, but check speed still limited by scanning speed, be up to 1km/h, cost is high.

Existing tunnel deformation detection mode is because of the limitation of its instrument or technology or with high costs or detection speed Degree is slower, cannot fully meet the status of tunnel operation.Must design a kind of advanced detection method, reduce detection at This, shortens the maintenance work time, reduces the influence to traffic, meets the needs that tunnel operation quickly detects.

As the development of unmanned air vehicle technique and image processing techniques has item for the deformation in tunnel operation maintenance Part develop it is a kind of based on photogrammetric unmanned plane tunnel deformation detection method so that following detection is more convenient and efficient.

Summary of the invention

In order to overcome the above-mentioned deficiencies of the prior art, the invention proposes a kind of based on photogrammetric unmanned plane tunnel change Shape detection method obtains tunnel cross-sectional image using the unmanned plane for carrying CCD camera and laser emitter, then by based on photography The algorithm of measurement handles acquired image, obtains the geometrical characteristic of subway tunnel cross sectional shape, so that analysis obtains Tunnel deformation situation discharges pre-warning signal in time, guarantees tunnel operation security.

The present invention is implemented as follows: it is a kind of based on photogrammetric unmanned plane tunnel deformation detection method, mainly by counting It is formed according to acquisition system, data processing and inversion system and result display systems, which is characterized in that the unmanned machine testing side Method are as follows:

1) the GPS start-stop coordinate of unmanned plane during flying is set according to Tunnel testing region；

2) the course line automatic cruising that unmanned plane is set according to step 1), automatic enabling vision system and laser radar auxiliary are patrolled Boat, and the laser beam emitting device by being mounted in unmanned plane issues the laser of the homogeneous and bright perpendicular to tunnel axis；

3) while unmanned plane issues laser according to step 2), pass through the image acquiring device pair being mounted on unmanned plane Tunnel laser profile carries out fast recording；

4) when the data acquisition of unmanned plane completion step 3), the tunnel laser profile of acquisition is fitted；

5) when the data processing for completing step 4), in conjunction with space positioning apparatus, by fitting section and a preceding testing result Or design profile compares, and shows deformation.

The GPS start-stop coordinate, including setting initial coordinate P₁, end coordinate P₂, unmanned plane height, speed, appearance State, course, guidance unmanned plane is by regulation axis flight.

The laser beam emitting device issues the laser perpendicular to tunnel axis homogeneous and bright, in unmanned plane during flying process In consistently form apparent tunnel contour lasing area.

The image acquiring device, including four CCD cameras, and angle between camera is adjusted, obtain best image weight Folded rate continuously takes pictures to laser section by setpoint frequency during unmanned plane during flying.

The cross section correct method is to pass through least square method fitted ellipse section.

Spatial position of the space positioning apparatus record unmanned plane in tunnel, can obtain with image acquiring device Image location information is corresponding, is quickly found out the specific location of tunnel deformation.

The present invention is based on photogrammetric unmanned plane tunnel deformation detection methods, when detection unmanned plane flies in tunnel When, deformation detection is periodically carried out to tunnel contour by the detection device of carrying, is realized to vcehicular tunnel, railway tunnel, subway The health detection in tunnel.

The beneficial effects of the present invention are: the present invention is with the obvious advantage compared with traditional detection mode.Due to using unmanned plane Carry laser emitter, the mode that CCD camera and photogrammetric technology combine obtains tunnel geometrical characteristic, this completely new non-to connect Touch measurement method.Detection speed can be greatly improved, guarantees detection accuracy, reduces testing cost, detection operates also more just It is prompt and efficient, it is able to satisfy the detection demand in different type tunnel.

Detailed description of the invention

Fig. 1 is that the present invention is based on photogrammetric unmanned plane tunnel deformation detection method structure composition figures

Fig. 2 work flow diagram of the present invention

Fig. 3 body schematic diagram of the present invention

In figure: 1. image acquiring devices, 2. laser beam emitting devices, 3. aircraft, 4. landing framves.

Specific embodiment

Such as Fig. 1, Fig. 2, shown in Fig. 3, the present invention is a kind of based on photogrammetric unmanned plane tunnel deformation detection method, packet Include data collection system, data processing and inversion system and result display systems.

Data collection system carries out data acquisition, including aircraft 3, power supply system by the detection device of UAV flight System, control system, image capturing system.

Power supply system is powered using the electric hybrid power supply of oil for the equipment of all carryings, to guarantee that equipment works normally.

Control system can set the initial coordinate P of unmanned plane using GPS and long-range control hybrid system, the system₁, knot Beam coordinate P₂, flying height, speed, posture and course, guidance unmanned plane reaches the mesh detected automatically by project flight is provided , and landing frame 4 is installed, facilitate landing.

As shown in figure 3, image capturing system includes image acquiring device 1, laser beam emitting device 2, space positioning apparatus.Swash Light emitting devices issues the laser perpendicular to tunnel longitudinal axis homogeneous and bright, and apparent tunnel is consistently formed during unmanned plane during flying Road profile lasing area.Image acquiring device carries out laser profile during unmanned plane during flying, through four CCD cameras quick Record, provides initial data for subsequent photogrammetric algorithm, to fit tunnel geometric cross section.Space positioning apparatus record Spatial position of the unmanned plane in tunnel, can be corresponding with the image location information that image acquiring device obtains, and establishes tunnel change Shaped position figure is quickly found out the specific location of tunnel deformation.

The data of data collection system acquisition back are carried out data processing, analysis by data processing and inversion system, and are passed To computer.Received image is pre-processed first, is converted to gray level image or binary system, school is carried out to image fault Just, and coordinate conversion is carried out, different image coordinates is transformed into the same coordinate system, geometrical characteristic matching is carried out, by most Small square law carries out ellipse fitting, obtains tunnel cross-section.The result data of acquisition, with design profile and last testing result Comparative analysis.End user can check the geometrical characteristic of tunnel contour in result display systems, understand the deformation feelings in tunnel Condition, and the specific location of tunnel deformation can be obtained.

Specific workflow of the present invention, specific workflow below are merely to illustrate the present invention rather than limit this hair Bright range:

1) the GPS start-stop coordinate P of unmanned plane during flying is set according to Tunnel testing region₁And P₂；

2) it flies according to the course line 1) planned, and the data collection system by being mounted on unmanned plane is adopted in real time Collect tunnel section profile information, obtains lasing area image；

3) when according to data acquisition 2) is carried out, the image information of acquisition is transmitted to computer and carries out data processing and analysis, The geometrical characteristic of tunnel cross-section is fitted with least square method；

4) after according to 3) Data Management Analysis is completed, result is output to result display systems, with design profile and upper one Secondary testing result comparison, can get information about tunnel deformation situation.

In order to obtain the image of covering whole cross section profile, it is equipped with 4 CCD cameras.Since every video camera has Independent video camera coordinates system, needs to carry out coordinate system conversion, so that result to be presented in unified coordinate system.

The initial data for being stored as color image, the section of subway tunnel is extracted using a series of images Processing Algorithm Profile.In this approach, for ease of calculation, color image is first be converted into gray level image.It, will after gray proces Treated, and image is converted to bianry image, extracts the geological information of subway tunnel.Because laser rays is bright and green, It is clear that the gray level and green component of pixel are to extract the important information of laser profile.It is colored using threshold segmentation method Image can be divided into bianry image, only include white (1) and black (0)

In pervious much researchs, the cross sectional shape of the subway tunnel of deformation is typically considered ellipse.Passing through will Cross section profile is fitted to ellipse, and the severity of deformation can be directly assessed by elliptical geometrical characteristic.

The highway or railway tunnel cross sectional shape of deformation are usually that top half elliptic adds lower rectangular.By by section Contour fitting can directly assess the severity of deformation at ellipse plus rectangle by the geometrical characteristic of superposition.

The severity of subsection deformation can by the fitted figure section in tunnel and obstructed cycle detection as a result, It is assessed with the difference of design shape.

Space positioning system obtains the spatial position of unmanned plane during flying, in coordinates computed, can carry out attitude rectification, More accurate tunnel geometrical characteristic figure is obtained, detection accuracy is improved.

By adopting the above technical scheme based on photogrammetric unmanned plane tunnel deformation detection method, unmanned plane can be passed through Realization detects entire tunnel deformation automatically, and can establish tunnel deformation library, in conjunction with detection data to tunnel deformation into Row security evaluation, and corresponding judgement is made to subsequent possible variation；The method that the above method replaces artificial detection, improves Speed is detected, testing cost is reduced, realizes intelligence, and do not influence the normal operation in tunnel.Disclosed in the present invention program Technological means be not limited only to technological means disclosed in above embodiment, further include by the above any combination of technical features institute The technical solution of composition.