CN112560707B - Mobile road surface detection method and system based on laser light source - Google Patents
Mobile road surface detection method and system based on laser light source Download PDFInfo
- Publication number
- CN112560707B CN112560707B CN202011506643.5A CN202011506643A CN112560707B CN 112560707 B CN112560707 B CN 112560707B CN 202011506643 A CN202011506643 A CN 202011506643A CN 112560707 B CN112560707 B CN 112560707B
- Authority
- CN
- China
- Prior art keywords
- road surface
- image
- color
- area
- pavement
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention provides a mobile pavement detection method and a mobile pavement detection system, wherein the method comprises the steps of irradiating a pavement by adopting a laser light source, and acquiring a color image under a visible spectrum and a gray image under an infrared spectrum, which are acquired by shooting the pavement by a multispectral camera; detecting the color image to obtain a road surface pollution area; detecting the gray level image to obtain a road surface pollution area and a road surface foreign matter area; extracting sub-areas corresponding to the road surface pollution area and the road surface foreign matter area from the color image respectively to form a 4-channel image; inputting the 4-channel image into a preset classification identification network; and acquiring the recognition result output by the classification recognition network. The method improves the coverage area of single detection.
Description
Technical Field
The invention belongs to the technical field of pavement foreign matter and damage detection, and particularly relates to a mobile pavement detection method and system based on a laser light source.
Background
In the maintenance work of highways, it is necessary to detect the damage of the road surface condition periodically. The road surface pathological change process is accurately mastered, and data support is provided for formulating a scientific maintenance plan. In the maintenance work of the airport pavement, the damage of the pavement is detected, foreign matters on the pavement are also detected, and the damage to an aircraft in the form of tire puncture and engine suction is avoided.
Mobile pavement inspection is a common technical framework. The detection by using the millimeter wave radar and the laser radar on the mobile platform is a common technical route, but the equipment cost is high, and the mobile detection speed is slow. Aiming at the problems, a technical route for the visible light/infrared imaging equipment and the structured light to cooperatively detect the road surface appears, so that the equipment cost is greatly reduced, and the detection speed is improved. However, the structured light is limited by technical conditions, and has the defect of small coverage area, so that the single-detection coverage capability is limited.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a mobile pavement detection method and system based on a laser light source, which can improve the single detection coverage area.
In a first aspect, a method for detecting a movable road surface based on a laser light source comprises the following steps:
irradiating a road surface by using a laser light source, and acquiring a color image under a visible spectrum and a gray image under an infrared spectrum, which are acquired by shooting the road surface by using a multi-spectral camera;
detecting the color image to obtain a road surface pollution area;
detecting the gray level image to obtain a road surface pollution area and a road surface foreign matter area;
extracting sub-areas corresponding to the road surface pollution area and the road surface foreign matter area from the color image respectively to form a 4-channel image;
inputting the 4-channel image into a preset classification identification network;
and acquiring the recognition result output by the classification recognition network.
Preferably, the detecting the color image to obtain the road surface pollution area specifically includes:
converting the color image from a three-channel mode of 'red + green + blue' to a three-channel mode of 'one-dimensional brightness + two-dimensional color'; extracting a one-dimensional brightness channel signal to generate a brightness value image;
extracting two-dimensional color channel signals to generate a first color value image and a second color value image;
and analyzing the brightness value image, the first color value image and the second color value image to obtain a road surface pollution area.
Preferably, the analyzing the luminance value image, the first color value image and the second color value image to obtain the road surface pollution area specifically includes:
calculating a brightness average value and a brightness standard sample difference according to the brightness values of all pixels in the brightness value image, and defining an area in which the brightness average value and the brightness standard sample difference fall within a preset brightness abnormal range as the road surface pollution area;
converting the luminance value image into a luminance binary image;
calculating a first color average value and a first color standard sample difference according to the first color value image, calculating a second color average value and a second color standard sample difference according to the second color value image, and defining an area in which the first color average value, the first color standard sample difference, the second color average value and the second color standard sample difference fall within a preset color abnormal range as the road surface pollution area;
converting the first color value image and the second color value image into a color binary image;
and performing logic OR calculation on each pixel in the brightness binary image and the color binary image, and defining an area with a pixel value of 1 in a calculation result as the road surface pollution area.
Preferably, the converting the luminance value image into the luminance binary image specifically includes:
converting the luminance value image B (x, y) into a luminance outlier image difB (x, y) according to the following equation;
wherein, b avg Is the color average value, b std Is the color standard sample difference;
and converting the brightness outlier image into the brightness binary image according to a preset brightness conversion threshold value.
Preferably, the converting the first color value image and the second color value image into the color binary image specifically includes:
converting the first color value image C1 (x, y) and the second color value image C2 (x, y) into a color outlier image difC (x, y) according to the following equation;
wherein, c1 avg Is said first color average value, c1 std Is a first color standard sample difference; c2 avg Is the second color average value, c2 std Is the second color standard sample difference;
and converting the color outlier image into the color binary image according to a preset color conversion threshold value.
Preferably, the detecting the grayscale image to obtain a road surface pollution area and a road surface foreign matter area specifically includes:
correcting the gray level image according to a preset attenuation model to obtain a corrected image; the attenuation model is I / (x, y) = I (x, y)/sin (alpha), wherein alpha is an included angle between a ground imaging point corresponding to the pixel and a laser light source connecting line and a ground plane, I (x, y) is a gray image, and I (x, y) is / (x, y) is a rectified image;
and analyzing the corrected image to obtain a road surface pollution area and a road surface foreign matter area.
Preferably, the analyzing the corrected image to obtain a road surface pollution area and a road surface foreign matter area specifically includes:
defining an area with a gray value lower than a preset minimum gray value in the corrected image as the road surface pollution area;
calculating a position correction parameter delta of a road surface pollution area according to the following formula;
delta＝htan(α)；
wherein h is the maximum difference value of pixels in the road surface pollution area in the x coordinate direction; alpha is an included angle between a connecting line of a ground imaging point corresponding to the pixel and the laser light source and the ground plane;
and calculating the foreign matter area of the road surface according to the road surface pollution area and the position correction parameter, wherein the x coordinate distribution of pixels in the foreign matter area of the road surface is the same as that of the pixels in the road surface pollution area, the upper limit of the y coordinate of the pixels in the foreign matter area of the road surface is the lower limit of the y coordinate of the pixels in the road surface pollution area, and the lower limit of the y coordinate of the pixels in the foreign matter area of the road surface is the sum of the upper limit of the y coordinate of the pixels in the road surface pollution area and the position correction parameter delta.
Preferably, the classification identification network is used for defining the target as the foreign matter on the road surface when the target which does not belong to the inherent structure of the road surface and is damaged is identified;
the training method of the classification recognition network comprises the following steps:
a: establishing a training sample library containing normal pavement samples, pavement damaged samples and pavement foreign matter samples;
b: initializing a classification identification network according to a normal pavement sample and a pavement damage sample;
c: inputting a normal pavement sample and a pavement foreign matter sample into a classification and identification network, recording a characteristic vector output by the classification and identification network, and calculating classification parameters according to a support vector machine method;
d: and D, replacing the classification decision layer parameters of the classification and recognition network with the obtained classification parameters, continuously inputting the classification and recognition network according to the normal pavement sample and the pavement damage sample, and repeatedly executing the step C to finish the training of the classification and recognition network.
Preferably, the classification identification network is specifically configured to:
obtaining the inherent structure of the road surface of the target and the similarity score of the damage;
judging whether the obtained similarity score is lower than a preset inherent structure and a damage threshold of the road surface, and if so, defining the target as a foreign matter of the road surface; if not, the target is defined as the category corresponding to the highest similarity score.
In a second aspect, a mobile pavement detection system based on a laser light source comprises a mobile platform, wherein a calculation control unit and a plurality of detection units electrically connected with the calculation control unit are arranged on the mobile platform;
the detection unit is provided with a multispectral camera and a laser light source, wherein the mounting height of the multispectral camera is greater than that of the laser light source, and the multispectral camera and the laser light source are mounted towards the front lower part in a downward overlooking manner; the visual field coverage range of the multispectral camera is the same as the irradiation range of the laser light source;
the calculation control unit is configured to execute the mobile pavement detection method according to the first aspect.
According to the technical scheme, the mobile pavement detection method and system based on the laser light source provided by the invention can improve the single detection coverage area.
Drawings
In order to more clearly illustrate the detailed description of the invention or the technical solutions in the prior art, the drawings used in the detailed description or the prior art description will be briefly described below. Throughout the drawings, like elements or portions are generally identified by like reference numerals. In the drawings, elements or portions are not necessarily drawn to scale.
Fig. 1 is a flowchart of a mobile pavement detection method according to an embodiment of the present invention.
Fig. 2 is a block diagram of a mobile pavement detection system according to a fifth embodiment of the present invention.
Detailed Description
Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and therefore are only examples, and the protection scope of the present invention is not limited thereby. It is to be noted that, unless otherwise specified, technical or scientific terms used herein shall have the ordinary meaning as understood by those skilled in the art to which the present invention belongs.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
As used in this specification and the appended claims, the term "if" may be interpreted contextually as "when", "upon" or "in response to a determination" or "in response to a detection". Similarly, the phrase "if it is determined" or "if a [ described condition or event ] is detected" may be interpreted contextually to mean "upon determining" or "in response to determining" or "upon detecting [ described condition or event ]" or "in response to detecting [ described condition or event ]".
The first embodiment is as follows:
a method for detecting a movable pavement based on a laser light source is disclosed, and referring to fig. 1, the method comprises the following steps:
s1: irradiating a road surface by using a laser light source, and acquiring a color image under a visible spectrum and a gray image under an infrared spectrum, which are acquired by shooting the road surface by using a multi-spectral camera;
s2: detecting the color image to obtain a road surface pollution area;
s3: detecting the gray level image to obtain a road surface pollution area and a road surface foreign matter area;
s4: extracting sub-areas corresponding to the road surface pollution area and the road surface foreign matter area from the color image respectively to form a 4-channel image;
s5: inputting the 4-channel image into a preset classification identification network;
specifically, the classification and identification network is used for identifying foreign matters and damages on the road surface of the input image. The foreign bodies on the road surface can be any objects, and have the characteristics of lack of generality and difficulty in exhaustion, so when the classification recognition network recognizes the inherent structure and damage of the road surface, the target which does not belong to the existing category in the classification recognition network is judged to be the foreign body.
S6: and acquiring the recognition result output by the classification recognition network.
According to the method, a laser light source illumination mode is used for replacing a structured light illumination mode in the prior art, visible light and infrared imaging equipment are combined for detecting foreign matters and damages on the road surface, and the single detection coverage area can be greatly increased. Meanwhile, the influence caused by the lack of the information of the structured light geometric pattern can be effectively compensated.
Example two:
second embodiment on the basis of the first embodiment, a method for detecting a road surface under a visible spectrum is defined.
The detecting the color image to obtain the road surface pollution area specifically comprises:
s21: converting the color image from a three-channel mode of 'red + green + blue' into a three-channel mode of 'one-dimensional brightness + two-dimensional color'; extracting a one-dimensional brightness channel signal to generate a brightness value image;
s22: extracting color channel signals of the two-dimensional color image to generate a first color value image and a second color value image;
s23: analyzing the brightness value image, the first color value image and the second color value image to obtain a pavement pollution area; the method specifically comprises the following steps:
s231: calculating a brightness average value b according to the brightness values of all pixels in the brightness value image avg And the standard sample difference b of brightness std Defining the mean value of the luminanceb avg And the standard sample difference b of brightness std The area falling into the preset abnormal brightness range is the road surface pollution area;
s232: converting the luminance value image into a luminance binary image, specifically comprising:
converting the luminance value image B (x, y) into a luminance outlier image difB (x, y) according to the following equation;
wherein (x, y) is the pixel coordinate, b avg The color mean value, b std Is the color standard sample difference;
and converting the brightness outlier image into the brightness binary image according to a preset brightness conversion threshold value.
S233: calculating a first color average c1 according to the first color value image avg And a first color standard sample difference c1 std Calculating a second color average c2 from the second color value image avg And a second color standard sample difference c2 std Defining a first color average value c1 avg First color standard sample difference c1 std Second color average value c2 avg And a second color standard sample difference c2 std The area falling into the preset color abnormal range is the road surface pollution area;
s234: converting the first color value image and the second color value image into a color binary image, specifically comprising:
converting the first color value image C1 (x, y) and the second color value image C2 (x, y) into a color outlier image difC (x, y) according to the following formula;
wherein, c1 avg Is the first color average value, c1 std Is the first color standard sample difference; c2 avg Is said second color average value, c2 std Is a second color markA quasi-sample difference;
converting the color outlier image into the color binary image according to a preset color conversion threshold
S235: and performing logic OR calculation on each pixel in the brightness binary image and the color binary image, and defining an area with a pixel value of 1 in a calculation result as the road surface pollution area.
Therefore, the method for detecting the road surface under the visible spectrum can effectively detect the foreign matters and damages on the road surface and the difference of the imaging color, the brightness and the texture of the inherent structure of the road surface.
For the sake of brief description, the method provided by the embodiment of the present invention may refer to the corresponding contents in the foregoing method embodiments.
Example three:
third embodiment on the basis of the above embodiments, a method for detecting a road surface under infrared spectroscopy is defined.
The detecting the gray level image to obtain a road surface pollution area and a road surface foreign matter area specifically comprises:
s31: correcting the gray level image I (x, y) according to a preset attenuation model to obtain a corrected image I / (x,y)；
Specifically, the attenuation model of the method is an attenuation model from near to far according to the laser irradiation intensity, and is used for correcting the irradiation light intensity unevenness. I is / (x, y) = I (x, y)/sin (alpha), wherein alpha is an included angle between a ground imaging point corresponding to the pixel and a connecting line of the laser light source and the ground plane.
S32: analyzing the corrected image to obtain a road surface pollution area and a road surface foreign matter area, and specifically comprising the following steps of:
s321: defining an area with a gray value lower than a preset minimum gray value in the corrected image as the road surface pollution area;
s322: calculating a position correction parameter delta of a road surface pollution area according to the following formula;
delta＝htan(α)；
wherein h is the maximum difference value of pixels in the road surface pollution area in the x coordinate direction; alpha is an included angle between a connecting line of a ground imaging point corresponding to the pixel and the laser light source and the ground plane;
s323: and calculating the foreign matter area of the road surface according to the road surface pollution area and the position correction parameter, wherein the x coordinate distribution of pixels in the foreign matter area of the road surface is the same as that of the pixels in the road surface pollution area, the upper limit of the y coordinate of the pixels in the foreign matter area of the road surface is the lower limit of the y coordinate of the pixels in the road surface pollution area, and the lower limit of the y coordinate of the pixels in the foreign matter area of the road surface is the sum of the upper limit of the y coordinate of the pixels in the road surface pollution area and the position correction parameter delta.
It is thus clear that the method for detecting a road surface in infrared spectroscopy can effectively detect the difference in shape between a flat road surface and undulation due to foreign matter and damage on the road surface.
For a brief description, the method provided by the embodiment of the present invention may refer to the corresponding content in the foregoing method embodiment.
Example four:
fourth embodiment, on the basis of the above embodiments, a training method and a recognition method of a classification recognition network are defined.
The classification identification network is used for defining a target as a foreign matter on the road surface when the road surface is identified to have the target which does not belong to the inherent structure of the road surface and is damaged;
the training method of the classification recognition network comprises the following steps:
a: establishing a training sample library containing a normal pavement sample, a pavement damaged sample and a pavement foreign matter sample;
b: initializing the classification recognition network according to the normal pavement sample and the pavement damage sample;
c: inputting a normal pavement sample and a pavement foreign matter sample into a classification and identification network, recording a characteristic vector output by the classification and identification network, and calculating classification parameters according to a support vector machine method;
d: and (4) replacing the classification decision layer parameters of the classification and recognition network with the obtained classification parameters, continuously inputting the classification and recognition network according to the normal pavement sample and the pavement damage sample, repeatedly executing the step C, and finishing the training of the classification and recognition network until the normal pavement sample, the pavement damage sample and the pavement foreign matter sample have better recognition accuracy.
Wherein the classification recognition network is specifically configured to:
obtaining the inherent structure of the road surface of the target and the similarity score of the damage;
judging whether the obtained similarity score is lower than a preset inherent structure and a damage threshold of the road surface, and if so, defining the target as a foreign matter of the road surface; if not, the target is defined as the category corresponding to the highest similarity score.
Therefore, the classification identification network identification method can identify the road surface foreign matters through identification and elimination, and solves the problems that the road surface foreign matters can be any objects, lack generality and are difficult to exhaust.
For the sake of brief description, the method provided by the embodiment of the present invention may refer to the corresponding contents in the foregoing method embodiments.
Example five:
a mobile pavement detection system based on a laser light source comprises a mobile platform, see figure 2, wherein the mobile platform is provided with a calculation control unit and a plurality of detection units electrically connected with the calculation control unit;
the detection unit is provided with a multispectral camera and a laser light source, wherein the mounting height of the multispectral camera is greater than that of the laser light source, and the multispectral camera and the laser light source are mounted towards the front lower part in a downward overlooking manner; the visual field coverage range of the multispectral camera is the same as the irradiation range of the laser light source;
the calculation control unit is used for executing the mobile pavement detection method.
In particular, mobile platforms include mobile robots, mobile vehicles, and the like mobile devices. The mobile platform comprises a plurality of detection units, a wireless communication unit, a positioning unit and a calculation control unit which can be carried on the mobile platform. One detection unit includes one multispectral camera (e.g., a dual-spectral camera) and one laser light source. The detection unit, the wireless communication unit and the positioning unit are all linked to the calculation control unit, and the calculation control unit controls the mobile platform to move.
When the multispectral camera and the laser light source are installed, the multispectral camera can be installed at a higher height (for example, more than 1.5 meters) and is overlooked towards the front and the bottom. The laser light source may be installed at a low height (e.g., below 0.2 m) to irradiate toward the front lower side. Therefore, the road surface area covered by the visual field of the multispectral camera is the coverage area of the detection unit, and the irradiation range of the laser light source completely covers the road surface area covered by the visual field of the multispectral camera.
The system utilizes a detection unit with various coverage areas to enable the mobile platform to have the belt-shaped coverage capability from the center of the front to the two sides of the front. If the coverage range of the detection unit needs to be adjusted, so that the coverage range is farther, a longer focal length, a smaller overlooking angle and a higher installation position can be used, or the beam width or the divergence angle of the laser light source and the irradiation angle are adjusted, so that the full-coverage irradiation of the field of view of the multispectral camera is realized.
Preferably, the mobile pavement detection system is provided with a high-speed detection mode and a careful detection mode;
the calculation control unit is also used for setting the mode of the mobile pavement detection system to be a high-speed detection mode when the mobile pavement detection system is started, and setting the mode of the mobile pavement detection system to be a careful detection mode when the pavement pollution area or the pavement foreign matter area is identified.
In particular, the system has a high speed detection mode and a careful detection mode.
The system first starts a high speed detection mode and the system sends a start operation signal to the calculation control unit through the wireless communication unit. And the mobile platform enters a polling mode and runs according to the set running track and speed. Meanwhile, the detection unit continuously transmits the acquired image signals to the calculation control unit, and the calculation control unit detects foreign matters on the road surface and road surface damage. The calculation control unit collects the positioning signals of the positioning unit and transmits the positioning signals and the images to the background in real time through the wireless communication unit.
Under a high-speed detection mode, if foreign matters on the road surface or damage of the road surface are detected, the system immediately exits from the high-speed detection mode, enters a careful detection mode, approaches to a target according to a positioning signal, and transmits an image and positioning information of the detected target to a background after careful detection and confirmation. If the mobile platform has a cleaning function, the cleaning function can be started to clean the pavement, and after the pavement is cleaned, the high-speed detection mode is recovered.
For a brief description of the system provided by the embodiment of the present invention, reference may be made to the corresponding contents in the foregoing method embodiments.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit and scope of the present invention, and they should be construed as being included in the following claims and description.
Claims (10)
1. A mobile pavement detection method based on a laser light source is characterized by comprising the following steps:
irradiating a road surface by using a laser light source, and acquiring a color image under a visible spectrum and a gray image under an infrared spectrum, which are acquired by shooting the road surface by using a multispectral camera;
detecting the color image to obtain a road surface pollution area;
detecting the gray level image to obtain a road surface pollution area and a road surface foreign matter area;
extracting sub-regions corresponding to the pavement pollution regions from the color images, and extracting sub-regions corresponding to the pavement foreign matter regions from the gray-scale images to form 4-channel images;
inputting the 4-channel image into a preset classification identification network; and acquiring the recognition result output by the classification recognition network.
2. The method for detecting a moving pavement based on a laser light source according to claim 1, wherein the step of detecting the color image to obtain a polluted area of the pavement specifically comprises:
converting the color image from a three-channel mode of 'red + green + blue' into a three-channel mode of 'one-dimensional brightness + two-dimensional color'; extracting a one-dimensional brightness channel signal to generate a brightness value image;
extracting two-dimensional color channel signals to generate a first color value image and a second color value image;
and analyzing the brightness value image, the first color value image and the second color value image to obtain a road surface pollution area.
3. The method of claim 2, wherein the analyzing the luminance value image, the first color value image, and the second color value image to obtain the road surface pollution area specifically comprises:
calculating a brightness average value and a brightness standard sample difference according to the brightness values of all pixels in the brightness value image, and defining an area in which the brightness average value and the brightness standard sample difference fall within a preset brightness abnormal range as the road surface pollution area;
converting the luminance value image into a luminance binary image;
calculating a first color average value and a first color standard sample difference according to the first color value image, calculating a second color average value and a second color standard sample difference according to the second color value image, and defining an area in which the first color average value, the first color standard sample difference, the second color average value and the second color standard sample difference fall within a preset color abnormal range as the road surface pollution area;
converting the first color value image and the second color value image into a color binary image;
and performing logic OR calculation on each pixel in the brightness binary image and the color binary image, and defining an area with a pixel value of 1 in a calculation result as the road surface pollution area.
4. The method of claim 3, wherein converting the luminance value image into a luminance binary image comprises:
converting the luminance value image B (x, y) into a luminance outlier image difB (x, y) according to the following equation;
wherein, b avg Is the average value of the brightness, b std Is the luminance standard sample difference;
and converting the brightness outlier image into the brightness binary image according to a preset brightness conversion threshold value.
5. The method of claim 3, wherein converting the first color value image and the second color value image into color binary images comprises:
converting the first color value image C1 (x, y) and the second color value image C2 (x, y) into a color outlier image difC (x, y) according to the following formula;
wherein, c1 avg Is the first color average value, c1 std Is the first color standard sample difference; c2 avg Is said second color average value, c2 std Is the second color standard sample difference;
and converting the color outlier image into the color binary image according to a preset color conversion threshold value.
6. The laser light source-based mobile pavement detection method according to claim 1, wherein the detecting the grayscale image to obtain a pavement pollution area and a pavement foreign matter area specifically comprises:
correcting the gray level image according to a preset attenuation model to obtain a corrected image; the attenuation model is I / (x, y) = I (x, y)/sin (alpha), wherein alpha is an included angle between a ground imaging point corresponding to a pixel and a laser light source connecting line and a ground plane, I (x, y) is a gray image, and I (x, y) is a gray scale image / (x, y) is a rectified image;
and analyzing the corrected image to obtain a road surface pollution area and a road surface foreign matter area.
7. The method according to claim 6, wherein the analyzing the corrected image to obtain the contaminated area and the foreign area comprises:
defining an area with a gray value lower than a preset minimum gray value in the corrected image as the road surface pollution area;
calculating a position correction parameter delta of a road surface pollution area according to the following formula;
delta＝h tan(α)；
wherein h is the maximum difference value of pixels in the road surface pollution area in the x coordinate direction; alpha is an included angle between a connecting line of a ground imaging point corresponding to the pixel and the laser light source and the ground plane;
and calculating the foreign matter area of the road surface according to the road surface pollution area and the position correction parameter, wherein the x coordinate distribution of pixels in the foreign matter area of the road surface is the same as that of the pixels in the road surface pollution area, the upper limit of the y coordinate of the pixels in the foreign matter area of the road surface is the lower limit of the y coordinate of the pixels in the road surface pollution area, and the lower limit of the y coordinate of the pixels in the foreign matter area of the road surface is the sum of the upper limit of the y coordinate of the pixels in the road surface pollution area and the position correction parameter delta.
8. The method of claim 1, wherein the mobile pavement detection method comprises a step of detecting the position of the mobile pavement with a laser source
The classification identification network is used for defining the target as the foreign matter on the road surface when the road surface is identified to have the target which does not belong to the inherent structure of the road surface and is damaged;
the training method of the classification recognition network comprises the following steps:
a: establishing a training sample library containing normal pavement samples, pavement damaged samples and pavement foreign matter samples;
b: initializing a classification identification network according to a normal pavement sample and a pavement damage sample;
c: inputting a normal road surface sample and a road surface foreign matter sample into a classification and identification network, recording a characteristic vector output by the classification and identification network, and calculating classification parameters according to a support vector machine method;
d: and D, replacing the classification decision layer parameters of the classification and recognition network with the obtained classification parameters, continuously inputting the classification and recognition network according to the normal pavement sample and the pavement damage sample, and repeatedly executing the step C to finish the training of the classification and recognition network.
9. The method of claim 8, wherein the mobile pavement inspection method based on a laser source,
the classification recognition network is specifically configured to:
obtaining the inherent structure of the road surface of the target and the similarity score of the damage;
judging whether the obtained similarity score is lower than a preset inherent structure and a damage threshold of the road surface, and if so, defining the target as a foreign matter of the road surface; if not, the target is defined as the category corresponding to the highest similarity score.
10. A mobile pavement detection system based on a laser light source is characterized by comprising a mobile platform, wherein the mobile platform is provided with a calculation control unit and a plurality of detection units electrically connected with the calculation control unit;
the detection unit is provided with a multispectral camera and a laser light source, wherein the mounting height of the multispectral camera is greater than that of the laser light source, and the multispectral camera and the laser light source are both mounted towards the front lower part in an overlooking manner; the visual field coverage range of the multispectral camera is the same as the irradiation range of the laser light source;
the computing control unit is used for executing the mobile pavement detection method of any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011506643.5A CN112560707B (en) | 2020-12-18 | 2020-12-18 | Mobile road surface detection method and system based on laser light source |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011506643.5A CN112560707B (en) | 2020-12-18 | 2020-12-18 | Mobile road surface detection method and system based on laser light source |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112560707A CN112560707A (en) | 2021-03-26 |
| CN112560707B true CN112560707B (en) | 2022-10-21 |
Family
ID=75030641
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011506643.5A Active CN112560707B (en) | 2020-12-18 | 2020-12-18 | Mobile road surface detection method and system based on laser light source |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112560707B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103576159A (en) * | 2013-11-14 | 2014-02-12 | 中国民用航空总局第二研究所 | Runway pavement detection device and method based on laser scanning technology |
| CN105222752A (en) * | 2015-09-29 | 2016-01-06 | 中国民用航空总局第二研究所 | A kind of portable road face pick-up unit and method using structured light |
| CN109870457A (en) * | 2019-02-14 | 2019-06-11 | 武汉武大卓越科技有限责任公司 | Track foreign matter detecting method and device |
| CN110133744A (en) * | 2018-02-09 | 2019-08-16 | 北京康力优蓝机器人科技有限公司 | A kind of mobile robot road surface road conditions detection method |
| CN111145161A (en) * | 2019-12-28 | 2020-05-12 | 北京工业大学 | Method for processing and identifying pavement crack digital image |
| CN111485475A (en) * | 2020-04-23 | 2020-08-04 | 陕西国防工业职业技术学院 | Pavement pit recognition method and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105157626B (en) * | 2015-09-29 | 2018-01-02 | 中国民用航空总局第二研究所 | A kind of fixed road face detection means and method using structure light |
| GB201711412D0 (en) * | 2016-12-30 | 2017-08-30 | Maxu Tech Inc | Early entry |
-
2020
- 2020-12-18 CN CN202011506643.5A patent/CN112560707B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103576159A (en) * | 2013-11-14 | 2014-02-12 | 中国民用航空总局第二研究所 | Runway pavement detection device and method based on laser scanning technology |
| CN105222752A (en) * | 2015-09-29 | 2016-01-06 | 中国民用航空总局第二研究所 | A kind of portable road face pick-up unit and method using structured light |
| CN110133744A (en) * | 2018-02-09 | 2019-08-16 | 北京康力优蓝机器人科技有限公司 | A kind of mobile robot road surface road conditions detection method |
| CN109870457A (en) * | 2019-02-14 | 2019-06-11 | 武汉武大卓越科技有限责任公司 | Track foreign matter detecting method and device |
| CN111145161A (en) * | 2019-12-28 | 2020-05-12 | 北京工业大学 | Method for processing and identifying pavement crack digital image |
| CN111485475A (en) * | 2020-04-23 | 2020-08-04 | 陕西国防工业职业技术学院 | Pavement pit recognition method and device |
Non-Patent Citations (5)
| Title |
|---|
| Detection and Quantification of Road Surface Damage using Digital Image Processing Techniques;Mohan N, Asst. Prof等;《International Journal of Engineering Research & Technology (IJERT)》;20170630;第6卷(第06期);185-188 * |
| Mobile laser scan data for road surface damage detection;Van Der Horst, B. B.等;《International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences》;20191231;1141-1148 * |
| 基于视频图像处理的城市道路污染检测与识别;任婷婷;《中国优秀硕士学位论文全文数据库_信息科技辑》;20160815;I138-656 * |
| 基于视频的机场跑道异物检测算法研究;周扬;《中国优秀硕士学位论文全文数据库_信息科技辑》;20150215;I138-828 * |
| 机场刚性道面破损的表观形貌分析及薄层修补研究;于磊等;《低温建筑技术》;20200128;第42卷(第01期);第131-134页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112560707A (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110261436B (en) | Rail fault detection method and system based on infrared thermal imaging and computer vision | |
| CN111079747B (en) | Railway wagon bogie side frame fracture fault image identification method | |
| CN102663760B (en) | Location and segmentation method for windshield area of vehicle in images | |
| US8670612B2 (en) | Environment recognition device and environment recognition method | |
| CN110033431B (en) | Non-contact detection device and detection method for detecting corrosion area on surface of steel bridge | |
| CN110175576A (en) | A kind of driving vehicle visible detection method of combination laser point cloud data | |
| CN102073846B (en) | Method for acquiring traffic information based on aerial images | |
| CN110211101A (en) | A kind of rail surface defect rapid detection system and method | |
| CN111091544B (en) | Method for detecting breakage fault of side integrated framework of railway wagon bogie | |
| CN104899554A (en) | Vehicle ranging method based on monocular vision | |
| CN105373135A (en) | Method and system for guiding airplane docking and identifying airplane type based on machine vision | |
| CN113822840A (en) | Vehicle bottom inspection system and method | |
| CN110954968B (en) | Airport runway foreign matter detection device and method | |
| CN110427979B (en) | Road water pit identification method based on K-Means clustering algorithm | |
| CN114578807B (en) | Unmanned target vehicle radar fusion active target detection and obstacle avoidance method | |
| WO2024108971A1 (en) | Agv system for vehicle chassis corrosion evaluation | |
| CN107341455A (en) | A kind of detection method and detection means to the region multiple features of exotic on night airfield runway road surface | |
| CN112508911A (en) | Rail joint touch net suspension support component crack detection system based on inspection robot and detection method thereof | |
| CN111598846A (en) | Rail defect detection method in tunnel based on YOLO | |
| CN114638835B (en) | Sleeper foreign matter detection method based on depth camera | |
| CN112883778A (en) | Road well lid height difference identification method and device based on computer vision | |
| CN112288682A (en) | Electric power equipment defect positioning method based on image registration | |
| CN116309407A (en) | Method for detecting abnormal state of railway contact net bolt | |
| CN114529555A (en) | Image recognition-based efficient cigarette box in-and-out detection method | |
| CN112560707B (en) | Mobile road surface detection method and system based on laser light source |
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 |