A method for detecting road construction flatness

Abstract

The application discloses a kind of methods for detecting road construction flatness, comprising: projecting structured light with specific shape, capturing road surface image;The captured road surface image is pretreated;Road surface features are extracted from pretreated image;The mode recognition is carried out to the extracted road surface feature, and the road flatness is calculated;Road flatness is analyzed, and visual result is output;Wherein, mode recognition includes: according to the extracted road surface feature parameter, design is used to distinguish different road flatness mode classifier;The extracted road surface feature is combined into a feature vector;According to feature vector, the road flatness is classified;Through cross validation and parameter adjustment, the classifier is optimized;The application improves detection accuracy and efficiency, reduces the need for manual intervention;Technical method is simple and easy to operate, and the complexity of equipment and algorithm is low, suitable for real-time detection and monitoring.

Description

Technical Field

[0001] This invention relates to a method for detecting the smoothness of road construction, belonging to the field of road construction technology. Background Technology

[0002] During road construction, asphalt and water-stabilized pavements typically employ a method of compaction followed by testing. However, if the pavement smoothness fails the smoothness test, the roller will struggle to repair the uneven surface as temperatures drop. Therefore, online smoothness detection is crucial for road finishing work. Smoothness process detection and control first involves detecting and analyzing the construction conditions and the environment of the pavement to be paved. Structured light detection and feature extraction technologies can determine the pavement's smoothness and construction scenario. This information clarifies the paver's location and interaction status. Next, based on the specific interaction scenario, logical reasoning is used to determine the compaction behavior using the reverse elevation control process for paving operations. Among multiple selectable behaviors, the optimal docking behavior is selected based on the paving task, screed elevation control, and the paver's real-time pose information to control smoothness docking. Finally, the docking control information is transmitted to the ground roller vehicle. By detecting ground information, the ground roller is guided to perform directional operation control. This method helps ensure the smoothness and quality of road construction. Summary of the Invention

[0003] The technical problem solved by this invention is that the compaction process in the prior art cannot detect the road surface smoothness and the operator compacts blindly.

[0004] To solve the above-mentioned technical problems, the present invention is implemented using the following technical solution.

[0005] On one hand, the present invention provides a method for detecting the smoothness of road construction, comprising:

[0006] Structured light is projected in a specific shape to capture images of the road surface;

[0007] The captured road surface image is preprocessed to obtain a preprocessed image;

[0008] Road surface features are extracted from the preprocessed image;

[0009] The extracted road surface features are subjected to pattern recognition to calculate the road smoothness;

[0010] Analyze the road smoothness and output the visualization results;

[0011] The pattern recognition includes:

[0012] Based on the extracted road surface feature parameters, a classifier is designed to distinguish different road smoothness patterns;

[0013] The extracted road surface features are combined into a feature vector;

[0014] Road smoothness is classified based on the feature vectors;

[0015] The classifier was optimized through cross-validation and parameter tuning.

[0016] This invention employs structured light detection technology, which can obtain high-precision topographic information of the road surface under test, improving detection accuracy and efficiency. The pattern recognition method realizes automated processing and analysis of road smoothness data, reducing the need for manual intervention. The technical method is simple and easy to implement, with low complexity of equipment and algorithms, making it suitable for real-time detection and monitoring.

[0017] The specific shape is one of the shapes of light stripes and spots emitted by structured light.

[0018] By processing the captured structured light stripes, the center line of the stripes can be extracted, thus improving the detection accuracy.

[0019] The preprocessing includes at least one of noise reduction, noise denoising, distortion correction, and image enhancement.

[0020] The noise reduction employs Gaussian filtering image processing technology, where a Gaussian function describes the image. A little bit The formula for pixel values ​​is as follows:

[0021]

[0022] in, On the original image pixel value of a point These are Gaussian filter parameters, representing the width of the Gaussian function.

[0023] The denoising employs a median filtering algorithm, including:

[0024] Use a moving window with an odd number of points to align the center of the window with a certain pixel position in the image;

[0025] Read the grayscale values ​​of the corresponding pixels in the window;

[0026] Select the grayscale value of the middle pixel in the grayscale sequence;

[0027] Assign the intermediate value to the pixel at the center of the window.

[0028] Median filtering can reduce noise in an image and improve image quality.

[0029] The distortion correction involves calculating image coordinates based on camera internal parameters and distortion coefficients, and then generating a corrected image using an interpolation algorithm.

[0030] Distortion correction can reduce image distortion caused by camera system or lens problems, such as radial and tangential distortion, making the shapes of objects in the image closer to reality.

[0031] The image enhancement includes contrast enhancement, edge enhancement, and adaptive histogram equalization.

[0032] Image enhancement can improve the visual quality of an image or facilitate subsequent processing and analysis, avoiding the over-enhancement problem that may be caused by global histogram equalization.

[0033] Extracting road surface features from the preprocessed image includes: using a local binary mode algorithm to extract local texture features from the preprocessed image to form feature parameters.

[0034] The calculation of road smoothness includes: calculating the standard deviation or root mean square deviation of the height difference;

[0035] The standard deviation of the height difference is calculated using the following formula:

[0036]

[0037] in, The number of samples in the dataset. For the first The height difference of each sample This is the average height difference of all samples;

[0038] The root mean square deviation of the height difference is calculated using the following formula:

[0039]

[0040] in, The number of samples in the dataset. For the first The height difference of each sample.

[0041] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for detecting the smoothness of road construction.

[0042] Beneficial effects

[0043] This invention utilizes structured light detection and feature extraction technologies to accurately determine the current smoothness of the road surface and the specific construction scenario, significantly improving the overall quality and efficiency of road construction. Through pattern recognition methods, it can adapt to different road types and environmental conditions, maintaining stable detection performance in complex environments. Through real-time detection and control, this invention provides a novel and intelligent solution for road construction. Attached Figure Description

[0044] Figure 1 This is a schematic flowchart of the method for detecting the smoothness of road construction as shown in Embodiment 1 of the present invention;

[0045] Figure 2 This is a schematic diagram of the calculation of the bilinear interpolation algorithm shown in Embodiment 1 of the present invention. Detailed Implementation

[0046] The present invention will be further described below with reference to specific embodiments.

[0047] Example 1:

[0048] like Figure 1 As shown, this embodiment provides a method for detecting the smoothness of road construction, including:

[0049] Structured light is projected in a specific shape to capture images of the road surface;

[0050] The captured road surface image is preprocessed to obtain a preprocessed image;

[0051] Extract road surface features from the preprocessed image;

[0052] Pattern recognition is performed on the extracted road surface features to calculate road smoothness;

[0053] Analyze road smoothness and output visualized results;

[0054] Pattern recognition includes:

[0055] Based on the extracted road surface feature parameters, a classifier is designed to distinguish different road smoothness patterns;

[0056] The extracted road surface features are combined into a feature vector;

[0057] Road smoothness is classified based on feature vectors;

[0058] The classifier was optimized through cross-validation and parameter tuning.

[0059] Specifically, structured light is laser light.

[0060] Specifically, autonomous vehicles integrate high-precision sensors such as LiDAR and high-definition cameras, which can capture the three-dimensional information, texture, and geometry of the road surface in real time. LiDAR scans the road surface with millimeter-level precision to obtain point cloud data; high-definition cameras capture high-definition images of the road surface.

[0061] Specifically, a camera or image sensor is used to capture structured light stripes on the road surface being tested; the captured structured light stripes are processed to extract the center line of the stripes; and the smoothness of the road surface being tested is calculated based on the center line.

[0062] Specifically, the results are output in numerical form, such as smoothness index and grade rating, to facilitate quantitative assessment of road quality. The results are also displayed in visual form, such as marking areas with different smoothness using different colors or textures, to provide a visual understanding of the road surface's smoothness.

[0063] The specific shape is one of the shapes of light stripes and spots emitted by structured light.

[0064] Preprocessing includes at least one of noise reduction, noise denoising, distortion correction, and image enhancement.

[0065] Noise reduction employs Gaussian filtering image processing techniques; the Gaussian function describes the image. A little bit The formula for pixel values ​​is as follows:

[0066]

[0067] in, On the original image pixel value of a point These are Gaussian filter parameters, representing the width of the Gaussian function.

[0068] Specifically, the point cloud data acquired by the lidar is filtered to remove noise and irrelevant information, while retaining the effective data related to road smoothness.

[0069] The noise reduction uses a median filtering algorithm, including:

[0070] Use a moving window with an odd number of points to align the center of the window with a certain pixel position in the image;

[0071] Read the grayscale values ​​of the corresponding pixels in the window;

[0072] Select the grayscale value of the middle pixel in the grayscale sequence;

[0073] Assign the intermediate value to the pixel at the center of the window.

[0074] Specifically, you can read the corresponding pixel grayscale values ​​under the template from small to large or from large to small.

[0075] Distortion correction involves calculating image coordinates based on camera internal parameters and distortion coefficients, and then using an interpolation algorithm to generate the corrected image.

[0076] Specifically, the camera's intrinsic parameters and distortion coefficients are obtained by photographing a calibration object of known shape and size (such as a checkerboard pattern). The image coordinates are then converted to camera coordinates (normalized camera coordinates). Based on the camera parameters and distortion coefficients, the distorted camera coordinates are calculated. These distorted camera coordinates are then converted back to image coordinates. An interpolation algorithm (such as bilinear interpolation) is used to interpolate the calculated point coordinates with the original image to obtain the corrected image.

[0077] Specifically, such as Figure 2 As shown, the bilinear interpolation algorithm is used to calculate the unknown function. At the green dot The coordinates and corresponding pixel values ​​of pixel values ​​Q11(x1,y1), Q12(x1,y2), Q21(x2,y1), and Q22(x2,y2) are given by the following formula:

[0078]

[0079]

[0080] exist Directional interpolation, and Substitute the values ​​and calculate the pixel values:

[0081]

[0082] in, This is the intermediate transition point for one-dimensional linear interpolation. This is the second intermediate transition point for one-dimensional linear interpolation.

[0083] Image enhancement includes contrast enhancement, edge enhancement, and adaptive histogram equalization.

[0084] Specifically, the images captured by the camera are enhanced to improve their contrast and clarity, facilitating subsequent feature extraction.

[0085] Extracting road surface features from preprocessed images includes: using a local binary mode algorithm to extract local texture features from the preprocessed images to form feature parameters.

[0086] Specifically, features related to road smoothness, such as elevation changes, slope, and curvature, are extracted from the processed image data. Computer vision techniques are used to extract features such as road surface texture, cracks, and potholes from the images.

[0087] Specifically, the Local Binary Pattern Algorithm (LBP) steps include: selecting a center pixel and its neighborhood; for each pixel in the image, selecting a pixel P as the center point; selecting a circular neighborhood (usually a 3x3 neighborhood, but can be expanded to a larger neighborhood, such as 5x5 or 7x7) with P as the center point, and N sampling points on the circular neighborhood (N is usually 8 for a 3x3 neighborhood); comparing the gray value of each sampling point with the gray value of the center point. If the gray value of the sampling point is greater than or equal to the gray value of the center point, the corresponding binary code of the sampling point is 1; otherwise, it is 0; arranging the binary codes obtained from the N sampling points in order (usually clockwise or counterclockwise) to obtain a binary number of length N; converting this binary number to a decimal number, which is the LBP value of the pixel; performing the above operations on each pixel in the image, and finally obtaining the local binary pattern feature map or local binary pattern histogram of the entire image.

[0088] Calculating road smoothness includes: calculating the standard deviation or root mean square deviation of the height difference;

[0089] The standard deviation of the height difference is calculated using the following formula:

[0090]

[0091] in, The number of samples in the dataset. For the first The height difference of each sample This is the average height difference of all samples;

[0092] The root mean square deviation of the height difference is calculated using the following formula:

[0093]

[0094] in, The number of samples in the dataset. For the first The height difference of each sample.

[0095] Specifically, standard deviation and root mean square deviation are both indicators of the dispersion of data center values. Standard deviation reflects the average dispersion of the data, while root mean square deviation reflects the overall dispersion of the data. When calculating the dispersion of height differences, either standard deviation or root mean square deviation can be used, depending on the application scenario and requirements.

[0096] Example 2:

[0097] This embodiment provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of a method for detecting the smoothness of road construction.

[0098] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A method for detecting the smoothness of road construction, characterized in that, include: Structured light is projected in a specific shape to capture images of the road surface; The captured road surface image is preprocessed to obtain a preprocessed image; Road surface features are extracted from the preprocessed image; The extracted road surface features are subjected to pattern recognition to calculate the road smoothness; Analyze the road smoothness and output the visualization results; The pattern recognition includes: Based on the extracted road surface feature parameters, a classifier is designed to distinguish different road smoothness patterns; The extracted road surface features are combined into a feature vector; Road smoothness is classified based on the feature vectors; The classifier was optimized through cross-validation and parameter tuning.

2. The method for detecting road construction smoothness according to claim 1, characterized in that, The specific shape is one of the shapes of light stripes and spots emitted by structured light.

3. The method for detecting the smoothness of road construction according to claim 1, characterized in that, The preprocessing includes at least one of noise reduction, noise denoising, distortion correction, and image enhancement.

4. The method for detecting the smoothness of road construction according to claim 3, characterized in that, The noise reduction employs Gaussian filtering image processing technology, where a Gaussian function describes the image. A little bit The formula for pixel values ​​is as follows: in, On the original image pixel value of a point These are Gaussian filter parameters, representing the width of the Gaussian function.

5. The method for detecting the smoothness of road construction according to claim 3, characterized in that, The denoising employs a median filtering algorithm, including: Use a moving window with an odd number of points to align the center of the window with a certain pixel position in the image; Read the grayscale values ​​of the corresponding pixels in the window; Select the grayscale value of the middle pixel in the grayscale sequence; Assign the intermediate value to the pixel at the center of the window.

6. The method for detecting the smoothness of road construction according to claim 3, characterized in that, The distortion correction involves calculating image coordinates based on camera internal parameters and distortion coefficients, and then generating a corrected image using an interpolation algorithm.

7. The method for detecting the smoothness of road construction according to claim 3, characterized in that, The image enhancement includes contrast enhancement, edge enhancement, and adaptive histogram equalization.

8. The method for detecting the smoothness of road construction according to claim 1, characterized in that, Extracting road surface features from the preprocessed image includes: using a local binary mode algorithm to extract local texture features from the preprocessed image to form feature parameters.

9. The method for detecting the smoothness of road construction according to claim 1, characterized in that, The calculation of road smoothness includes: calculating the standard deviation or root mean square deviation of the height difference; The standard deviation of the height difference is calculated using the following formula: in, The number of samples in the dataset. For the first The height difference of each sample This is the average height difference of all samples; The root mean square deviation of the height difference is calculated using the following formula: in, The number of samples in the dataset. For the first The height difference of each sample.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method for detecting the smoothness of road construction as described in any one of claims 1 to 9.

Images