CN118348015B - Self-movable subway tunnel structure apparent disease detection equipment and method - Google Patents
Self-movable subway tunnel structure apparent disease detection equipment and method Download PDFInfo
- Publication number
- CN118348015B CN118348015B CN202410611058.3A CN202410611058A CN118348015B CN 118348015 B CN118348015 B CN 118348015B CN 202410611058 A CN202410611058 A CN 202410611058A CN 118348015 B CN118348015 B CN 118348015B
- Authority
- CN
- China
- Prior art keywords
- image
- point cloud
- self
- camera
- array camera
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
- G01S17/06—Systems determining position data of a target
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/88—Lidar systems specially adapted for specific applications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8861—Determining coordinates of flaws
- G01N2021/8864—Mapping zones of defects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Analytical Chemistry (AREA)
- Pathology (AREA)
- Immunology (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Electromagnetism (AREA)
- Biochemistry (AREA)
- Computer Networks & Wireless Communication (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention provides a self-moving type subway tunnel structure apparent defect detection device and method, which are based on a wide area shooting area array camera, splice, process and explain acquired images, combine mobile measurement precise dead reckoning, can realize high-precision and rapid detection of multiple defects in the whole area of a subway tunnel, and aim to solve the problems of low manual detection speed, low defect identification dependence on traditional experience, low accuracy and low intelligent level of the defect detection device.
Description
Technical Field
The invention belongs to the technical field of tunnel detection, and particularly relates to self-moving type subway tunnel structure apparent defect detection equipment and method, in particular to self-moving type subway tunnel structure apparent defect detection equipment and method based on a wide-area shooting area array camera.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
With the rapid increase of economy and the rapid promotion of the urban mass transit, urban rail transit is rapidly developed. Because the tunnel environment is extremely complex, the geological environment is complex and is often influenced by factors such as operation vibration, construction disturbance, extreme weather and the like, the damage of the underground engineering structure is extremely easy to cause.
With the extension of service time, the operation mileage is increased year by year, the number of diseases is increased year by year, and the operation tunnel faces massive detection demands, however, the traditional detection method mainly uses manual detection, has low efficiency, difficult data backtracking and difficult accuracy guarantee, so that development of intelligent detection equipment for tunnel diseases is needed to meet the increasing disease detection demands.
Disclosure of Invention
The invention aims to solve the problems, and provides self-moving type subway tunnel structure apparent disease detection equipment and a method, wherein the invention is based on a wide area shooting area array camera, performs splicing, processing and interpretation on acquired images, combines mobile measurement precision dead reckoning, can realize high-precision and rapid detection on multiple diseases of a subway tunnel universe, and aims to solve the problems of low manual detection speed, low disease identification dependence on traditional experience, low accuracy and low intelligent level of the disease detection equipment.
According to some embodiments, the present invention employs the following technical solutions:
the utility model provides a from portable subway tunnel structure apparent disease detection equipment, includes from travelling car, wide area shooting area array camera, laser scanning system, control system and equips the platform, wherein:
the self-moving trolley is characterized in that a travelling mechanism is arranged at the bottom of the self-moving trolley and driven by a driving mechanism, and the travelling mechanism is used for moving bidirectionally/unidirectionally along a tunnel track and stopping at a set measuring point;
An equipment platform is arranged on the self-moving trolley, and the wide area shooting area array camera is rotatably arranged on the equipment platform;
The wide-area shooting area array camera is provided with a bracket in a semicircular arc form, and a plurality of area array cameras are sequentially arranged on the outer arc of the bracket;
The laser scanning system is used for acquiring point cloud data in the tunnel;
The control system is used for conducting dead reckoning according to information acquired by the displacement sensor and the mileage encoder, receiving images and point cloud data of each measuring point position, conducting panoramic stitching on each image according to each image acquired by the wide-area shooting area array camera, analyzing the stitched images to identify disease cracks, combining the dead reckoning and the point cloud data, fusing to generate a three-dimensional point cloud scene, mapping and registering the point cloud data and the image information, and achieving positioning of the disease cracks.
As an alternative implementation mode, the two ends of the self-moving trolley are provided with a trolley head, the trolley head is provided with a searchlighting system, and the trolley head is provided with a camera system and a radar system.
As an alternative implementation mode, the equipment platform is arranged at the middle position of the self-moving trolley, the lower end of the platform is provided with a lifting mechanism, and the platform is provided with a carrying space.
As an alternative implementation mode, a fixed guide rail is arranged on the equipment platform, a rotating device is movably arranged on the fixed guide rail, the rotating device has a certain degree of rotation freedom, the position of the rotating device on the fixed guide rail is changeable, a carrying device is arranged on the rotating device, and the support is arranged on the carrying device.
As an alternative implementation mode, the travelling mechanism is a wheel hub, the wheel hubs are arranged on two sides of the self-moving trolley, the wheel hubs travel along the track direction of the tunnel, the two sides of the wheel hubs are provided with liquid-gas mixing shock absorbers, and the wheel hubs are driven by a wheel hub motor;
and the walking mechanism is also provided with an inclination sensor for detecting the inclination angle of the self-moving trolley.
As an alternative implementation manner, a surface state measurer and a light supplementing light source are arranged in each surface of the surface array camera, and the surface array camera is provided with a lens automatic focusing module.
The working method of the apparent defect detection equipment based on the self-moving subway tunnel structure comprises the following steps of:
when the self-moving trolley moves to a preset measuring point position, a wide-area shooting area array camera and a laser scanning system are started at the same time, and image data and point cloud data of a corresponding scene at the same moment are obtained;
According to the information obtained by the displacement sensor and the mileage coder, dead reckoning is carried out, and the dead reckoning and the point cloud data are combined to generate a three-dimensional point cloud scene in a fusion way;
According to each image obtained by the wide-area shooting area array camera, carrying out panoramic stitching on each image, analyzing the stitched images, and identifying disease cracks;
And mapping and registering the point cloud data and the image information to realize the positioning of the disease cracks.
As an alternative embodiment, the specific process of dead reckoning includes:
acquiring coordinate information of a pre-configured measuring point;
Based on coordinate information of the measuring points, the current position acquired by the displacement sensor and the mileage encoder, forward dead reckoning and backward dead reckoning are carried out, and an initial dead reckoning position is determined;
combining the initial aerial survey position with the three-dimensional point cloud scene of the tunnel;
And correcting the initial aerial survey position by using actual sensor data, and fusing the corrected aerial survey position information with the three-dimensional point cloud scene again to form a final true three-dimensional point cloud scene.
As an alternative embodiment, the specific process of mapping and registering the point cloud data and the image information includes:
establishing a space projection relation between a laser scanning system and an area array camera, shooting a plurality of images in an environment with known geometric characteristics, scanning for a plurality of times, and calculating external parameters and internal parameters by using the acquired data;
Converting the point cloud and the image from respective three-dimensional coordinate systems to a unified two-dimensional plane by using a spatial projection relation obtained by calibration, and mapping the point cloud to the two-dimensional plane through an orthographic projection algorithm to form a point cloud image;
converting the point cloud image into a gray image on a two-dimensional plane, and accurately registering the point cloud gray image and the image by using an image processing technology;
and identifying the fine defects on the tunnel surface based on the point cloud gray level map and the registered image.
Alternatively, the process of acquiring the image by the wide-area shooting area array camera comprises the following steps:
Correcting the focal length and visual angle information of each camera lens to ensure that the position relation of each camera in space is consistent, and adjusting the parameters of each camera lens by knowing the size and distance of an object to enable the imaging results of each camera lens to correspond to each other;
Adopting a distributed automatic focusing algorithm to unify focusing results of all cameras;
And fusing imaging results of different camera lenses, and forming a complete image according to the information weight of each pixel.
As an alternative embodiment, the specific process of intelligent interpretation of the disease image comprises:
acquiring the existing images covering different types of diseases;
the image is subjected to a pre-processing step,
Extracting features including crack size, texture, shape, or edge information;
training a machine learning model using the extracted features;
and processing the acquired new image by using the trained machine learning model.
Compared with the prior art, the invention has the beneficial effects that:
According to the invention, by adopting the self-moving trolley with the multifunctional fixed equipment platform, different equipment can be automatically and quickly converted and carried, and the endurance mileage is further improved. The self-moving trolley is provided with a multifunctional self-moving carrying platform, has the characteristics of platform universalization, function modularization and interface standardization, has high modularization and automation, and can carry and rapidly switch different devices.
Aiming at the difficult problem of limited detection range, the invention utilizes the wide area shooting area array camera module in a semicircular arc form and is provided with the lens automatic focusing module, thereby realizing high-speed real-time synchronous sampling and disease scanning of tunnels with a certain angle and achieving optimal space utilization and lightweight design.
The planar state measurer and the light supplementing light source are arranged in the planar array camera measuring plane, so that the best resolution and imaging view angle of the single camera under different motion scenes are realized, the acquired image data can be ensured to be clear, frames are not lost, and the acquisition of high-definition images under the high-speed travelling condition is realized.
The wide-area shooting area array camera is arranged on the platform through the rotating mechanism, the position of the rotating mechanism is movable, the shooting position and the shooting angle can be flexibly adjusted according to the condition of a measuring point, the tilting sensor is arranged, and the tilting angle of the self-moving trolley is detected, so that the stability of the trolley when an image is shot is ensured, or the tilting angle is compensated by adjusting the gesture of the wide-area shooting area array camera.
The invention can realize the one-key automatic matching focusing of a plurality of lenses by utilizing the control system, and can ensure that the image acquisition is hardly interfered by illumination by utilizing the light supplement, thereby effectively improving the accuracy of later image splicing and image identification.
In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention.
FIG. 1 is a schematic structural diagram of apparent defect detection equipment for a self-moving subway tunnel structure;
fig. 2 is a schematic diagram of the structure of a wide-area photographing area camera.
Wherein, 1, a self-moving trolley, 2, a searchlighting system, 3, a self-moving trolley travelling system, 4, a camera system and a radar system, 5, an embedded control panel, 6, a multifunctional fixed equipment platform, 7, a front-back moving guide rail, 8, carrying device 9, area camera, 10, shape surface state measurer, 11, camera module mounting bracket, 12, rotating device, 13, rotating device fixed guide rail, 14 is laser scanning system.
Detailed Description
The invention will be further described with reference to the drawings and examples.
It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present invention. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
Embodiments of the application and features of the embodiments may be combined with each other without conflict.
Example 1
As shown in fig. 1, a self-moving subway tunnel structure apparent disease detection device based on a wide area photographing area array camera includes:
From travelling car 1, but two directions remove, advance wheel hub through the automobile body both sides and advance along tunnel track direction, the wheel hub inboard is equipped with displacement sensor, inclination sensor and mileage encoder, and the wheel hub both sides have hung the liquid-gas mixture bumper shock absorber, absorption impact that can be better improves the stability of traveling, has guaranteed the quick high-accuracy collection of image acquisition.
The two locomotive fronts of the self-moving trolley are both provided with the searchlighting system 2, an advanced attitude stabilizing and controlling system is carried, and the operation range is greatly widened under various severe conditions such as haze days and the like without being influenced by light for observation and use.
The camera system and the radar system 4 are positioned on the front side above the head of the self-moving trolley 1 and are used for detecting road conditions of a traveling road section and transmitting signals to a hub motor of a traveling hub.
The embedded control panel 5 is positioned at the front side in the unmanned cabin of the self-moving trolley, and is internally provided with a high-precision inertial navigation system, a wireless transmission and remote control system.
The multifunctional fixed equipment platform 6 is positioned between two heads of the self-moving trolley, and can be lifted by one key and matched with different equipment.
The wide-area shooting area array camera device is installed and fixed on the carrying device 8 in a semicircular arc shape, a rotating device 12 and a front-back moving guide rail 7 are arranged below the carrying device 8, and the rotating device 12 is movably arranged on the front-back moving guide rail 7, so that disease scanning of more than 240 degrees of tunnels can be realized.
Each area array camera 9 is provided with an electric focusing module for adjusting focal length, focusing and aperture, so that the adaptability of complex environments of equipment is improved.
Each area array camera 9 is also provided with a physical surface state measurer 10 to deal with the physical surface state measurement in the complex environment and discover diseases as soon as possible, and an LED light supplementing light source is arranged to uniformly and effectively illuminate the tunnel wall on the premise of not influencing the driving safety, so that the camera can still capture a sufficiently clear image in the high-speed driving process and ensure the shooting quality in the dim environment.
The camera module mounting bracket 11 is fixed to the mounting device 8 by a U-shaped anchor ear.
In the embodiment, the wide-area shooting area array camera adopts the high-density stroboscopic light source and the high-precision electronic control lens, so that high-definition image acquisition under the high-speed travelling condition can be realized, and the image acquisition is ensured to be hardly interfered by illumination through one-key automatic matching focusing of a plurality of lenses.
The system also comprises a processor or a control system, wherein the processor or the control system is used for conducting dead reckoning according to the information acquired by the displacement sensor and the mileage encoder, receiving images and point cloud data of each measuring point position, conducting panoramic stitching on each image according to each image acquired by the wide-area shooting area array camera, analyzing the stitched images to identify disease cracks, combining the dead reckoning and the point cloud data, fusing to generate a three-dimensional point cloud scene, mapping and registering the point cloud data and the image information, and realizing the positioning of the disease cracks.
In the embodiment, the laser scanning system can be arranged at one side of the headstock, the other side of the headstock is also reserved with a mounting port, the headstock can be adjusted and carried on the other side according to actual needs, the precision and the resolution of the laser scanner can be set through special control software embedded in the embedded control panel, the embedded control panel can perform scanning setting of the laser scanning system, and the setting is completed after the required resolution and precision are selected.
The laser scanning system consists of a long-distance laser scanner and a short-distance laser scanner, wherein the long-distance laser scanner is suitable for a large-size tunnel, and the short-distance laser scanner is suitable for a smaller tunnel or a situation requiring higher details.
The laser scanning system selected by the embodiment has a 360-degree all-round panoramic scanning function, has higher scanning speed, has the highest speed of 120 ten thousand points per second, and can rapidly and accurately capture the tunnel contour in the moving process of the vehicle. The laser scanning system has the accuracy of 0.020 mm and the highest resolution of 0.010 mm, and can accurately acquire the detailed information and complete three-dimensional data of the surface of a complex object.
Example two
The working method of the self-moving subway tunnel structure apparent defect detection device according to the first embodiment comprises the following steps:
when the self-moving trolley moves to a preset measuring point position, a wide-area shooting area array camera and a laser scanning system are started at the same time, and image data and point cloud data of a corresponding scene at the same moment are obtained;
According to the information obtained by the displacement sensor and the mileage coder, dead reckoning is carried out, and the dead reckoning and the point cloud data are combined to generate a three-dimensional point cloud scene in a fusion way;
According to each image obtained by the wide-area shooting area array camera, carrying out panoramic stitching on each image, analyzing the stitched images, and identifying disease cracks;
And mapping and registering the point cloud data and the image information to realize the positioning of the disease cracks.
The specific procedure is described below.
First, focusing of a plurality of lenses of a wide-area photographing area array camera and stitching of collected images. In some embodiments, panorama stitching software is configured, so that photos shot by each camera can be stitched into a panorama, the stitched images are analyzed in real time through the installed intelligent video analysis software, and wall cracks are identified through an algorithm.
In the embodiment, the distance information of the shooting object is detected by using a sensor on a camera or other optical equipment, and then the optimal focusing point is calculated by a built-in algorithm, so that the purpose of clear imaging is achieved. In a multiple lens system, each lens may have its own sensor and focusing module for independently performing an autofocus function.
The specific process comprises the following steps:
(1) Lens correction and synchronization in a system of a plurality of lenses, information such as a focal length, a viewing angle, etc. of each lens is first corrected to ensure that their positional relationship in space is uniform. By knowing the size and distance of the object, the parameters of each lens are adjusted so that their imaging results can correspond to each other.
(2) And an automatic focusing algorithm, namely adopting a distributed automatic focusing algorithm. Each lens performs independent automatic focusing according to own sensor reading, and then the results of the lenses are unified through an automatic focusing mechanism.
(3) And (3) image fusion and matching, namely fusing imaging results of different lenses after focusing the lenses, and forming a complete and high-quality image according to the information weight of each pixel.
(4) And (3) real-time feedback and adjustment, namely setting an image quality evaluation real-time feedback mechanism in the automatic matching focusing technology, and judging which focusing point has the highest imaging quality through an image definition evaluation function so as to adjust a focusing strategy according to the actual imaging effect.
In the disease identification process according to the spliced images, the existing detection frame or image identification network can be adopted to identify the disease. The method comprises the following specific steps:
(1) Image preprocessing, including resizing, cropping, denoising, contrast enhancement, etc., to facilitate subsequent processing and analysis. The diseased region may be highlighted by adjusting the brightness and contrast of the image or a filter may be used to remove noise from the image.
(2) Feature extraction-following image preprocessing, a feature extraction stage follows. The purpose of this stage is to extract disease-related features from the image for subsequent identification and classification.
The feature extraction of the embodiment utilizes a deep learning technology to automatically extract features, and mainly comprises the steps of data collection and preprocessing, data labeling, proper deep learning model selection, model training, verification and evaluation, feature extraction, post-processing and analysis.
A. And data collection and preprocessing, namely firstly collecting a large amount of image data of tunnel defects, and then preprocessing the images to improve the accuracy and efficiency of model training and testing. The preprocessing steps may include image enhancement (e.g., rotation, scaling, cropping), denoising, contrast adjustment, brightness adjustment, image size normalization, etc.
B. Labeling data-labels that may be bounding boxes, polygons, or pixel-level labels.
C. And selecting a proper deep learning model according to the specific requirements of the task. The common models are Convolutional Neural Networks (CNNs) such as AlexNet, VGGNet, resNet, and for the image segmentation problem, models such as U-Net, FCN, mask R-CNN, and the like can be selected.
D. Model training-training a selected deep learning model using the annotated data set. This process involves adjusting model parameters to minimize the loss function and thereby improve the predictive power of the model. Super-parameter tuning may also be required during the training process, such as learning rate, batch size, selection of optimizers, etc.
E. Verification and evaluation the performance of the model was evaluated on a separate test set to verify its generalization ability. Common evaluation indexes include accuracy, recall, F1 score, ioU (cross-over ratio), and the like. Depending on the evaluation, a return may be required to adjust the model structure or training strategy.
F. feature extraction once the model is trained and validated, it can be used to extract features of new tunnel defect images. This is typically achieved by forward propagation, i.e., inputting images into a trained model, outputting a representation of the features learned by the model.
G. Post-processing and analysis the extracted features typically require further processing and analysis to obtain information about the type, location, severity, etc. of tunnel defects. This may involve machine learning methods such as cluster analysis, anomaly detection, regression analysis, etc.
(3) Disease identification using machine learning or deep learning models to identify disease in an image. Convolutional Neural Networks (CNNs) may be used to identify tunnel cracks or other types of defects. Model training typically requires a large amount of annotation data to learn how to distinguish between normal and abnormal image features.
(4) Result verification and evaluation the recognition results are verified and evaluated to ensure accuracy and reliability of recognition, and a separate test dataset is used to evaluate the performance of the model.
The embodiment can be combined with an Inertial Navigation System (INS), laser scanning and other sensor data, so that dynamic precision measurement is realized under the condition of no GNSS signal, and the accumulated error of high-frequency position and attitude information provided by the inertial navigation system is effectively reduced through a real-time correction and data fusion technology, and the precision of integral measurement is improved. The absolute accuracy of the tunnel dynamic measurement is improved to 3 mm.
The specific dead reckoning process comprises the following steps:
(1) Coordinate information of a predefined control point (or measuring point) is acquired. The control point may be a specific position in the tunnel, or may be a key point such as an entrance and an exit of the tunnel.
(2) Forward/reverse dead reckoning initial POS dead reckoning methods predict future locations from known location information and current sensor data. Forward dead reckoning refers to predicting the position of the self-moving trolley in the tunnel from a known starting point, and reverse dead reckoning refers to reversely predicting the position of the self-moving trolley from a known ending point. By both methods, initial train position information, i.e., initial POS, can be obtained.
(3) And (3) fusing the initial three-dimensional point cloud scene, namely combining the initial POS with the three-dimensional point cloud scene of the tunnel. A point cloud is a data set that contains a large number of three-dimensional coordinate points representing the surface of an object. By fusing the initial POS and the point cloud scene, a more real three-dimensional tunnel model can be created.
(4) Track correction acquires accurate POS information by correcting the initial POS using actual sensor data to obtain more accurate location information.
(5) And (3) forming a fused true three-dimensional point cloud scene, namely fusing the corrected accurate POS information with the three-dimensional point cloud scene again to form a true fused true three-dimensional point cloud scene reflecting the current position of the self-moving trolley.
Aiming at the problem that the noise of the tunnel structure surface influences the reliability of the tunnel structure analysis, the embodiment utilizes a tunnel point cloud denoising method based on wavelet analysis, and in the denoising aspect, wavelet transformation can decompose a signal or an image into a series of sub-signals or sub-images on the scale and the azimuth, wherein the sub-signals or the sub-images comprise different frequency components of an original signal or image.
By appropriate processing of these sub-signals or sub-images, noise is effectively removed, preserving the information of the useful signals or images. The processes of scale segmentation, noise extraction, filtering and the like of the point cloud data of the tunnel section are realized through wavelet transformation, the problem of interference of tunnel auxiliary facilities and surface defects on measurement accuracy is solved, and the automatic extraction of parameters such as tunnel convergence, deformation, dislocation, intrusion and the like is realized.
Of course, other denoising methods may be employed.
The embodiment also provides a method for converting the point cloud and the image into a two-dimensional tunnel point cloud image and an image through space orthographic projection, and then realizing pixel-level fusion registration of the point cloud gray image and the image through calibration in order to improve the accuracy and precision of disease detection.
The method specifically comprises the following steps:
(1) And the time synchronization is to ensure that the time stamps of the data collected by the laser scanning system and the area array camera are the same so as to ensure that the point cloud and the image data correspond to the scene at the same moment.
(2) And (3) space calibration, namely establishing a space projection relation between the laser scanning system and the area array camera. Multiple images are taken and scanned multiple times in an environment of known geometry, from which external parameters (e.g., rotation and translation matrices) and internal parameters (e.g., focal length and distortion coefficients of the camera) are calculated.
(3) And (3) spatial orthographic projection, namely converting the point cloud and the image from respective three-dimensional coordinate systems to a unified two-dimensional plane by using a spatial projection relation obtained by calibration. The point cloud is mapped onto a two-dimensional plane through an orthographic projection algorithm to form a point cloud image, and the original two-dimensional characteristics of the image are maintained to form an image.
(4) Pixel level registration, namely, on a two-dimensional plane, pixel points of a point cloud image and an image can be directly corresponding. Since point cloud data typically contains gray scale information, point cloud images can be converted into gray scale images. And then, accurate registration of the point cloud gray level map and the image map is realized through image processing technologies such as feature matching and transformation model fitting.
(5) And after finishing the registration, filtering, denoising, enhancing and the like can be performed on the fused data so as to improve the quality of the final three-dimensional model.
By the method, high precision of point cloud data and rich color information of image data can be fully utilized, and powerful data support is provided for detection, evaluation and maintenance of tunnels.
Through converting the point cloud data into a gray level map or a depth map, the fine defects on the surface of the tunnel can be analyzed and identified by utilizing computer vision and machine learning technologies, the problems of fine defects (0.3 mm cracks, 5mm falling blocks, water seepage) on the surface of the tunnel and the like are effectively solved, and the difficult problem of rapid positioning analysis and extraction of the railway tunnel defects is solved.
The embodiment can realize high-precision and rapid detection of the whole domain multiple diseases of the subway tunnel, and aims to solve the problems of low manual detection speed, low accuracy rate and low intelligent level of disease detection equipment due to the fact that disease identification depends on traditional experience.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which do not require the inventive effort by those skilled in the art, are intended to be included within the scope of the present invention.
Claims (9)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410611058.3A CN118348015B (en) | 2024-05-16 | 2024-05-16 | Self-movable subway tunnel structure apparent disease detection equipment and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410611058.3A CN118348015B (en) | 2024-05-16 | 2024-05-16 | Self-movable subway tunnel structure apparent disease detection equipment and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118348015A CN118348015A (en) | 2024-07-16 |
| CN118348015B true CN118348015B (en) | 2025-02-14 |
Family
ID=91815791
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410611058.3A Active CN118348015B (en) | 2024-05-16 | 2024-05-16 | Self-movable subway tunnel structure apparent disease detection equipment and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118348015B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118612547A (en) * | 2024-08-07 | 2024-09-06 | 云南云路工程检测有限公司 | A camera automatic focusing method based on depth estimation |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114002704A (en) * | 2021-10-21 | 2022-02-01 | 深圳市明日系统集成有限公司 | A lidar monitoring method, device and medium for bridge and tunnel |
| CN115937408A (en) * | 2022-10-14 | 2023-04-07 | 山东大学 | Rail transit tunnel apparent disease high-precision identification and real scene model construction device and method |
| CN116183626A (en) * | 2023-02-15 | 2023-05-30 | 武汉理工大学 | A device and method for detecting bridge pier defects based on 3D visualization |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9731744B2 (en) * | 2015-09-04 | 2017-08-15 | Gatekeeper Systems, Inc. | Estimating motion of wheeled carts |
| CN107316325B (en) * | 2017-06-07 | 2020-09-22 | 华南理工大学 | Airborne laser point cloud and image registration fusion method based on image registration |
| CN207496669U (en) * | 2017-07-19 | 2018-06-15 | 广东技术师范学院 | A kind of vehicle safety travel intelligence auxiliary device for including road conditions monitoring |
| CN108053367B (en) * | 2017-12-08 | 2021-04-20 | 北京信息科技大学 | A 3D point cloud stitching and fusion method based on RGB-D feature matching |
| CN109131334A (en) * | 2018-08-30 | 2019-01-04 | 天津布尔科技有限公司 | By driving prior-warning device to the auxiliary of road conditions before video-splicing technology analysis vehicle |
| CN111536940A (en) * | 2020-05-14 | 2020-08-14 | 南京航空航天大学 | A tunnel over-under-excavation detection system and its detection method |
| CN214149430U (en) * | 2021-03-04 | 2021-09-07 | 山东交通学院 | Tunnel safety monitoring system applying mobile three-dimensional laser measurement system |
-
2024
- 2024-05-16 CN CN202410611058.3A patent/CN118348015B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114002704A (en) * | 2021-10-21 | 2022-02-01 | 深圳市明日系统集成有限公司 | A lidar monitoring method, device and medium for bridge and tunnel |
| CN115937408A (en) * | 2022-10-14 | 2023-04-07 | 山东大学 | Rail transit tunnel apparent disease high-precision identification and real scene model construction device and method |
| CN116183626A (en) * | 2023-02-15 | 2023-05-30 | 武汉理工大学 | A device and method for detecting bridge pier defects based on 3D visualization |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118348015A (en) | 2024-07-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11035958B2 (en) | Systems and methods for correcting a high-definition map based on detection of obstructing objects | |
| JP6904614B2 (en) | Object detection device, prediction model creation device, object detection method and program | |
| US11061122B2 (en) | High-definition map acquisition system | |
| CN112505065A (en) | Method for detecting surface defects of large part by indoor unmanned aerial vehicle | |
| JP2020525809A (en) | System and method for updating high resolution maps based on binocular images | |
| CN115439621B (en) | Three-dimensional map reconstruction and target detection method for underground coal mine inspection robot | |
| CN116403186B (en) | FPN Swin Transformer and Pointnet ++ based automatic driving three-dimensional target detection method | |
| CN110443819B (en) | Method and device for detecting track of monorail train | |
| CN110766760B (en) | Method, device, equipment and storage medium for camera calibration | |
| CN112883778B (en) | A method and device for identifying height difference of road manhole covers based on computer vision | |
| CN113034439B (en) | High-speed railway sound barrier defect detection method and device | |
| CN114092551A (en) | A smart car based on panoramic vision and lidar fusion SLAM system | |
| CN118348015B (en) | Self-movable subway tunnel structure apparent disease detection equipment and method | |
| CN115857040A (en) | Dynamic visual detection device and method for foreign matters on locomotive roof | |
| CN110766761A (en) | Method, device, equipment and storage medium for camera calibration | |
| Liu et al. | Lightweight defect detection equipment for road tunnels | |
| CN119516503A (en) | Road inspection method and system based on data fusion and edge computing | |
| CN119618189A (en) | High-precision map construction method for 'sky-air-ground' integrated off-road environment | |
| JP2009052907A (en) | Foreign object detection system | |
| JP3465531B2 (en) | Object recognition method and apparatus | |
| CN120539045A (en) | Method for detecting friction coefficient of airport runway based on close-range camera technology | |
| CN117037007B (en) | An aerial photography method and device for inspecting road lighting uniformity | |
| CN119339352B (en) | Road traffic sign board missing detection method based on vehicle-mounted front view image | |
| CN119296457B (en) | Projection vehicle lighting method and device based on image semantic segmentation and target recognition | |
| CN116433641A (en) | Fuel tank opening point pose estimation method based on refueling robot |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |