Geotechnical engineering investigation method for linear engineering

Abstract

The invention discloses a geotechnical engineering investigation method for linear engineering, which comprises the following steps of: dividing an inspection area into a plurality of working sections, inserting a sensor into each working section, and shooting geotechnical investigation image data of the inspection area by utilizing an inspection robot according to a set inspection route; the sensor receives the geotechnical investigation image data of each inspection working section, classifies the geotechnical investigation image data, and automatically identifies the classified categories in the image data shot by the inspection robot; and carrying out data fusion on the shooting data of the inspection robot in each working section to form complete image data. According to the invention, the investigation work of automatic inspection, automatic classification and automatic drawing is realized, the working intensity of workers is reduced, and the working efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of geotechnical engineering investigation, and in particular to a geotechnical engineering investigation method for linear engineering. Background Art

[0002] Linear engineering refers to comprehensive projects such as railways, highways, oil and gas pipelines, channels, pipelines, urban integrated pipe networks, transmission lines and cableways. During the linear engineering survey process, the target site must first be patrolled and observed, and then the target site must be divided into several work sections. The soil quality of each work section must be tested and classified. Finally, a manual work map is required to aggregate the maps of multiple work sections into a complete map. In the actual work process, the above steps all need to be completed manually, especially drawing the complete map, which is very tedious and greatly reduces work efficiency. Summary of the Invention

[0003] The purpose of the present invention is to overcome the shortcomings and defects of the existing technology and provide a geotechnical engineering survey method for linear engineering to realize the survey work of automatic inspection, automatic classification and automatic drawing, reduce the workload of staff and improve work efficiency.

[0004] The present invention is achieved in that:

[0005] A geotechnical engineering investigation method for a linear project comprises the following steps:

[0006] The inspection area is divided into several working sections, sensors are inserted into each working section, and the inspection robot is used to capture geotechnical survey image data of the inspection area according to the set inspection route;

[0007] The sensor receives geotechnical survey image data of each patrol section, classifies the geotechnical survey image data, and automatically identifies the classified categories in the image data taken by the patrol robot;

[0008] The data captured by the inspection robots in each work section are fused to form complete image data.

[0009] Among them, clustering algorithm is used to classify geotechnical survey image data and determine the geological composition of rock and soil.

[0010] Wherein, the clustering algorithm at least adopts the kmeans clustering algorithm.

[0011] The data captured by the inspection robots in each inspection area are fused to form complete image data. The image registration technology is used to stitch the images with overlapping areas into a complete image, including:

[0012] First, based on the feature extraction algorithm, feature data containing overlapping areas are extracted, and image registration is performed based on the extracted feature data to determine the corresponding point relationship of the overlapping areas of the input images;

[0013] An improved optimization algorithm based on dynamic programming is used to quickly stitch images, and then a weighted average algorithm is used to fuse data to form complete image data.

[0014] The method of using an improved optimization algorithm based on dynamic programming to rapidly stitch images includes:

[0015] Mark each pixel point in the first row of the overlapping area as the starting point of a feature data and calculate the intensity value of each point;

[0016] Extend the search for each starting point intensity value to the next row, and continue searching until the last row;

[0017] Add the intensity values ​​of each current point and the three adjacent pixels in the next row, find the pixel in the next row corresponding to the minimum intensity value, and this pixel is the expansion direction found among the three pixels. Modify the intensity value of the feature data to the minimum intensity value, and modify the current point to the pixel in the next row corresponding to the minimum intensity value, and then gradually search downward until the last row.

[0018] Among them, in the data fusion using the weighted average algorithm, the calculation formula of the weighted smoothing algorithm is as follows:

[0019] Where w1 and w2 are the weights of the corresponding pixels in the overlapping area of ​​the images to be stitched, respectively, satisfying w1+w2=1, 0<w1, w2<1.

[0020] Among them, the data fusion of the inspection robot shooting data of each working section to form complete image data is completed by a sensor of multiple working sections, and the sensor for image stitching is configured to receive the marked image data of the sensors of other working sections for image stitching.

[0021] Among them, the data shooting data of the inspection robots in each working section are fused to form complete image data, which is performed by an independent stitching processor for image stitching. The sensors of all working sections send the marked image data to a separate stitching processor for image stitching for image stitching processing.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] The present invention adopts the inspection robot to realize the automatic geotechnical engineering survey of linear engineering, which is more efficient than the geotechnical engineering survey of manual linear engineering.

[0024] The present invention uses a clustering algorithm to identify the data on the sensor, which can quickly and accurately realize the automatic classification of geotechnical data, and realize the rapid and accurate detection and classification of the soil quality of each working section;

[0025] The present invention uses data fusion to splice data to form inspection area map data, reducing the difficulty in drawing. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a schematic flow chart of the geotechnical engineering investigation method for linear engineering of the present invention; DETAILED DESCRIPTION

[0027] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0028] See also Figure 1 As shown, in an exemplary embodiment of the present application, a geotechnical engineering investigation method for a linear project is provided, comprising the following steps:

[0029] The inspection area is divided into several working sections, sensors are inserted into each working section, and the inspection robot is used to capture geotechnical survey image data of the inspection area according to the set inspection route;

[0030] The sensor receives geotechnical survey image data of each patrol section, classifies the geotechnical survey image data, and automatically identifies the classified categories in the image data taken by the patrol robot;

[0031] The data captured by the inspection robots in each work section are fused to form complete image data.

[0032] In an embodiment of the present application, there are multiple sensors, each of which is used to process the image data of the working section taken by the inspection robot. Based on the taken image data, the geological composition of the rock and soil in the working section is analyzed, the classification is determined, and then the classification is automatically identified in the image data taken by the inspection robot.

[0033] In an embodiment of the present application, after processing, analyzing and classifying the geotechnical survey image data through the sensors of each working section, the classification of the geological components of the geotechnical components of each working section in the inspection area can be obtained, and the data is automatically marked in the image data taken by the inspection robot. Finally, after splicing the image data of each working section taken by the inspection robot, a complete image with the classification of the geological components of the geotechnical components can be obtained.

[0034] Through the above technologies, the present invention realizes the survey work of automatic inspection, automatic classification, and automatic mapping, reduces the workload of staff and improves work efficiency.

[0035] Exemplarily, in the present application, the inspection robot may be an inspection aircraft or a humanoid inspection robot. The inspection robot is equipped with a high-definition camera or a high-definition image acquisition device, which is used to inspect each working section in its inspection area, collect images of the relevant working sections that can be used to subsequently classify the rock and soil components, and send the collected images of each working section wirelessly to the sensors of the corresponding working sections for processing and classification.

[0036] Exemplarily, the sensor is a data processing device having a data processor, including one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement a method for clustering analysis and processing of geotechnical survey image data, and implement clustering analysis of the received data through built-in relevant data processing algorithms, such as the clustering algorithm described below.

[0037] Illustratively, in an embodiment of the present application, the sensor receives geotechnical survey image data from each patrol section, classifies the geotechnical survey image data, and automatically identifies the classified categories in the image data taken by the patrol robot. In the next step of image data stitching, the image stitching can be completed by a sensor of multiple work sections. The sensor for image stitching is configured to receive the marked image data from sensors of other work sections for image stitching, or the sensors of all work sections send the marked image data to a separate processor for image stitching for image stitching processing.

[0038] The above-mentioned sensor for image stitching or a separate processor for image stitching includes one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement a method for image stitching processing of geotechnical exploration image data.

[0039] Illustratively, in an embodiment of the present application, a clustering algorithm is used to classify geotechnical survey image data to determine the geological composition of the rock and soil.

[0040] Exemplarily, in an embodiment of the present application, the clustering algorithm at least adopts the kmeans clustering algorithm.

[0041] For example, in the embodiment of the present application, the kmeans clustering algorithm transforms the geotechnical feature data into x (m), put the current survey data into the database to obtain the training sample set {x (1) 、x (2) ...x (m)}, then set k classes and the centroid of each class as μ1, μ2, ..., μ k , repeat the following process until convergence {

[0042] For each sample i, calculate the class it should belong to;

[0043]

[0044] For each class j, recalculate the centroid of the class;

[0045]

[0046] x (i) Belongs to the training sample set {x (1) 、x (2) ...x (m)}, c (1) represents the class that is closest to the sample i among the k classes, c (i) The value of is one from 1 to k. The centroid μ j Represents the expected value of the center point of samples belonging to the same class.

[0047] From the above calculation, we can get C (m) , that is, the class that is closest to the sample m among the k classes.

[0048] For example, in the embodiment of the present application, the data captured by the inspection robots in each inspection area are fused to form complete image data, which is achieved by using image registration technology to stitch the images containing overlapping areas into a complete image, including:

[0049] First, based on the feature extraction algorithm, feature data containing overlapping areas are extracted, and image registration is performed based on the extracted feature data to determine the corresponding point relationship of the overlapping area of ​​the input image; an improved optimization algorithm based on dynamic programming is used to quickly stitch images, and then a weighted average algorithm is used to fuse data to form complete image data.

[0050] Exemplarily, in an embodiment of the present application, the feature extraction algorithm can adopt an improved SURF algorithm based on a circular field. First, feature points are extracted through the Hessian matrix, and then the circular field of the feature points is used for feature description. A descriptor is established for each feature point using the Haar wavelet response. At the same time, the normalized grayscale difference and second-order gradient in the field are calculated to form a new feature descriptor. The minimum Euclidean distance criterion is used for feature point matching, and the RANSAC algorithm is used to eliminate mismatched points to further improve the matching accuracy of the algorithm. Other feature extraction algorithms can also be used, but are not limited to this.

[0051] For example, in the embodiment of the present application, the method of using an improved optimization algorithm based on dynamic programming to rapidly stitch images includes:

[0052] Mark each pixel point in the first row of the overlapping area as the starting point of a feature data and calculate the intensity value of each point; expand the search for each starting point intensity value to the next row and continue searching until the last row; add the intensity values ​​of each current point and the three pixel points adjacent to this point in the next row, and find the pixel point in the next row corresponding to the minimum intensity value. This pixel point is the extension direction found among the three pixels, and modify the intensity value of the feature data to the minimum intensity value. Modify the current point to the pixel point in the next row corresponding to the minimum intensity value, and then gradually search downward until the last row.

[0053] For example, in the embodiment of the present application, in the data fusion using the weighted average algorithm, the calculation formula of the weighted smoothing algorithm is as follows:

[0054]

[0055] Where w1 and w2 are the weights of the corresponding pixels in the overlapping area of ​​the images to be stitched, respectively, satisfying w1+w2=1, 0<w1, w2<1.

[0056] In the above weighted smoothing algorithm, the formula defines a function f(x,y), which determines its value based on the area where the point (x,y) is located: when the point (x,y) belongs only to image f1 (that is, (x,y)∈f1), f(x,y)=f1(x,y), that is, the pixel value of image f1 at that point is directly taken. When the point (x,y) belongs only to image f2 (that is, (x,y)∈f2), f(x,y)=f2(x,y), that is, the pixel value of image f2 at that point is directly taken. When the point (x,y) is in the overlapping area of ​​images f1 and f2 (that is, (x,y)∈(f1∩f2)), f(x,y)=w1f1(x,y)+w2f2(x,y).

[0057] w1 and w2 are weight coefficients, corresponding to the weights of the pixels in the overlapping area of ​​images f1 and f2, respectively. In the overlapping area, the values ​​of the corresponding pixels of the two images are fused by weighted averaging. The distribution of weights determines the contribution of the two images in the overlapping part to the final fusion result.

[0058] This weighted smoothing algorithm is commonly used in image stitching. When stitching multiple images into a complete image, the overlapping areas between the different images need to be fused to avoid stitching artifacts and make the stitched image appear more natural and continuous. By properly setting weights, the effects of the different images in the overlapping areas can be balanced, improving the quality of the stitched image.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] The present invention adopts the inspection robot to realize the automatic geotechnical engineering survey of linear engineering, which is more efficient than the geotechnical engineering survey of manual linear engineering.

[0061] The present invention uses a clustering algorithm to identify the data on the sensor, which can quickly and accurately realize the automatic classification of geotechnical data, and realize the rapid and accurate detection and classification of the soil quality of each working section;

[0062] The present invention uses data fusion to splice data to form inspection area map data, reducing the difficulty in drawing.

[0063] The basic principles, main features and advantages of the present invention are shown and described above. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments and that the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention.

[0064] The embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the claims are therefore intended to be embraced therein.

[0065] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A geotechnical engineering investigation method for a hydraulic linear project, characterized in that: The following steps are involved: The inspection area is divided into several working sections, sensors are inserted into each working section, and the inspection robot is used to capture geotechnical survey image data of the inspection area according to the set inspection route; The sensor receives geotechnical survey image data of each working section, classifies the geotechnical survey image data, and automatically identifies the classified categories in the image data taken by the inspection robot; The data captured by the inspection robots in each work section are fused to form complete image data.

2. The geotechnical engineering investigation method for hydraulic linear projects according to claim 1, characterized in that: Clustering algorithm is used to classify geotechnical investigation image data and determine the geological composition of rock and soil.

3. The geotechnical engineering investigation method for water conservancy linear projects according to claim 2 is characterized in that. The clustering algorithm at least adopts the kmeans clustering algorithm.

4. The geotechnical engineering investigation method for hydraulic linear engineering according to claim 1, characterized in that: The data captured by the inspection robot in each working section is fused to form complete image data. The image registration technology is used to stitch the images with overlapping areas into a whole image, including: First, based on the feature extraction algorithm, feature data containing overlapping areas are extracted, and image registration is performed based on the extracted feature data to determine the corresponding point relationship of the overlapping areas of the input images; An improved optimization algorithm based on dynamic programming is used to quickly stitch images, and then a weighted average algorithm is used to fuse data to form complete image data.

5. The geotechnical engineering investigation method for hydraulic linear projects according to claim 4, characterized in that: The improved optimization algorithm based on dynamic programming is used to quickly stitch images, including: Mark each pixel point in the first row of the overlapping area as the starting point of a feature data and calculate the intensity value of each point; Extend the search for each starting point intensity value to the next row, and continue searching until the last row; Add the intensity values ​​of each current point and the three adjacent pixels in the next row, find the pixel in the next row corresponding to the minimum intensity value, and this pixel is the expansion direction found among the three pixels. Modify the intensity value of the feature data to the minimum intensity value, and modify the current point to the pixel in the next row corresponding to the minimum intensity value, and then gradually search downward until the last row.

6. The geotechnical engineering investigation method for a hydraulic linear project according to claim 4, characterized in that: In the data fusion using the weighted average algorithm, the calculation formula of the weighted smoothing algorithm is as follows: Where w1 and w2 are the weights of the corresponding pixels in the overlapping area of ​​the images to be stitched, respectively, satisfying w1+w2=1, 0<w1, w2<1.

7. The geotechnical engineering investigation method for a hydraulic linear project according to claim 1, characterized in that: The data fusion of the inspection robot shooting data of each working section to form complete image data is completed by a sensor of multiple working sections. The sensor for image stitching is configured to receive the marked image data of the sensors of other working sections for image stitching.

8. The geotechnical engineering investigation method for a hydraulic linear project according to claim 1, characterized in that: The data shooting data of the inspection robots in each working section are fused to form complete image data, which is performed by an independent stitching processor for image stitching. The sensors of all working sections send the marked image data to a separate stitching processor for image stitching for image stitching processing.

Images