CN115830029A - Spring soil detection method based on computer vision - Google Patents
Spring soil detection method based on computer vision Download PDFInfo
- Publication number
- CN115830029A CN115830029A CN202310139076.1A CN202310139076A CN115830029A CN 115830029 A CN115830029 A CN 115830029A CN 202310139076 A CN202310139076 A CN 202310139076A CN 115830029 A CN115830029 A CN 115830029A
- Authority
- CN
- China
- Prior art keywords
- soil
- pixel
- value
- depth image
- area
- 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.)
- Granted
Links
Images
Landscapes
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention relates to the technical field of image data processing, in particular to a spring soil detection method based on computer vision. The invention improves the detection accuracy of the spring soil and reduces the detection cost at the same time.
Description
Technical Field
The invention relates to the technical field of image data processing, in particular to a spring soil detection method based on computer vision.
Background
The channel engineering is an important component of the capital construction engineering, and the channel engineering itself includes but is not limited to earth excavation and filling processes, highway engineering and roadbed engineering, and most of the construction processes of the engineering need to be carried out on the basis of soil body, while in the construction process, due to the change of the water content of the soil body, an engineering disease which becomes spring soil can occur, and if the problem of the spring soil can not be handled in time, the construction quality and the construction process can be seriously affected. When the spring soil problem is comparatively serious, can follow the soil surface characteristic and directly discern the spring soil, and when the spring soil problem was more slight, can only detect discernment spring soil through the water content to soil. However, because the construction site is relatively large, it is not practical to detect the construction site by using a method for detecting the water content of the soil, so in order to improve the detection efficiency of the spring soil, the spring soil detection of the construction site is generally realized by using a method for training a neural network in the prior art.
In practice, the inventors found that the above prior art has the following disadvantages:
in the prior art, data collected by construction equipment in the construction process is generally used as training data to train a neural network, and the detection of the spring soil in a construction site is realized according to the trained neural network, but the method for training the neural network needs a large amount of historical data to ensure the recognition accuracy, the cost is high, the accuracy of the neural network is seriously influenced when the historical data has error data, and the accuracy of correspondingly detecting the spring soil is insufficient. Therefore, the method for detecting the spring soil by training the neural network in the prior art is high in cost and inaccurate.
Disclosure of Invention
In order to solve the technical problems of high cost and inaccuracy of the method for detecting the spring soil by training the neural network in the prior art, the invention aims to provide a spring soil detection method based on computer vision, and the adopted technical scheme is as follows:
the invention provides a spring soil detection method based on computer vision, which comprises the following steps:
acquiring an original depth image before load and a load depth image after load of each soil compacting area, and acquiring a pressure change height map according to the difference between the original depth image and the load depth image of each soil compacting area;
obtaining a concave area and a convex area according to the pixel value in the original depth image; obtaining the molding compression degree of each soil compacting region in the pressure change height map according to the pixel value distribution characteristics of the pixel points in the concave region and the convex region;
obtaining a shaping flatness according to the pixel value variation characteristic difference between the original depth image and the load depth image, and obtaining a shaping deformation degree according to the shaping compression degree and the shaping flatness degree of each soil compacting area;
obtaining a pixel point characteristic value sequence in the load depth image according to the movement direction of the soil compactor; acquiring the ground elasticity significance of each soil compaction area according to the element value distribution trend characteristics in the pixel point characteristic value sequence;
and obtaining the spring soil significance of each compacted soil area according to the plastic deformation degree and the ground elasticity significance of each compacted soil area, and completing spring soil detection according to the spring soil significance.
Further, the method for acquiring the concave area and the convex area of the pressure change height map comprises the following steps:
and dividing the pixel values of all the pixel points in the original depth image by adopting an OTSU maximum inter-class variance method to obtain a division threshold, wherein the division threshold is a depth threshold, the pixel points corresponding to the pixel values in the original depth image which are greater than the depth threshold are recorded as concave pixel points, the pixel points corresponding to the pixel values in the original depth image which are less than the depth threshold are recorded as convex pixel points, the regions formed by all the concave pixel points are recorded as concave regions, and the regions formed by all the convex pixel points are recorded as convex regions.
Further, the method for obtaining the plastic compression degree comprises the following steps:
in the pressure change height map, calculating the pixel value mean value of all pixel points in the concave area and recording the pixel value mean value as a concave area characteristic value, calculating the pixel value mean value of all pixel points in the convex area and recording the pixel value mean value as a convex area characteristic value, calculating the difference value between the depth threshold value and the convex area characteristic value and recording the difference value as a convex difference value, calculating the difference value between the concave area characteristic value and the depth threshold value and recording the difference value as a concave difference value, and taking the ratio of the convex difference value to the concave difference value as the weight of the convex area characteristic value; and adding the result of multiplying the weight by the characteristic value of the convex area to the characteristic value of the concave area to obtain the plastic compression degree of the corresponding soil compacting area.
Further, the method for acquiring the pixel value variation characteristic comprises the following steps:
optionally selecting one of the original depth image and the loaded depth image as a target image;
taking each pixel point of the target image as a central pixel point, acquiring a gray level co-occurrence matrix according to the pixel value of the pixel point in a preset neighborhood range of the central pixel point, calculating the contrast corresponding to the gray level co-occurrence matrix, and recording the accumulated sum of the contrast corresponding to all the pixel points in the target image as a pixel value change characteristic.
Further, the method for obtaining the shaping flatness comprises the following steps:
and calculating the difference between the pixel value change characteristics of the pressure change height image and the load depth image in the same soil compaction area, recording the difference as the flatness difference, and recording the ratio of the flatness difference to the pixel value change characteristics of the load depth image as the shaping flatness of the corresponding soil compaction area.
Further, the method for acquiring the pixel point characteristic value sequence comprises the following steps:
in the load depth image, taking the direction opposite to the movement direction of the soil compactor as the stress characteristic direction;
and counting the pixel value median of each row of pixels in the load depth image, recording the pixel value median as a pixel characteristic value, and arranging the pixel characteristic values in a stress characteristic direction as a sequence to obtain a pixel characteristic value sequence.
Further, the method for acquiring the element value distribution trend feature in the pixel point feature value sequence comprises the following steps:
obtaining a hurst index by adopting a polymerization variance method for the pixel point characteristic value sequence, and recording the hurst index of the pixel point characteristic value sequence as an autocorrelation characteristic value of the load depth image;
calculating element difference values between every two adjacent elements in the pixel point characteristic value sequence, recording the element difference values as pixel point difference characteristic values, forming a sequence by using the pixel point difference characteristic values as elements according to the sequence of the elements in the pixel point characteristic value sequence, and recording the sequence as the pixel point difference characteristic value sequence;
calculating the standard deviation among various elements in the pixel point difference characteristic value sequence, recording the standard deviation as the characteristic value difference standard deviation of the load depth image, carrying out abnormal data detection on the pixel point difference characteristic value sequence and counting the number of abnormal values;
and recording the autocorrelation characteristic value, the pixel point difference characteristic value, the characteristic value difference standard deviation and the abnormal value quantity of the load depth image as the distribution trend characteristic.
Further, the method for acquiring the ground elasticity significance comprises the following steps:
acquiring the load depth image of the target compacted soil area, recording the product of the characteristic value difference standard deviation and the abnormal value quantity of the load depth image as an abnormal condition characteristic value, calculating the ratio of the autocorrelation characteristic value of the load depth image to the abnormal condition characteristic value plus a preset constant coefficient, recording the ratio as an inclined characteristic value, recording the sum of the difference characteristic values of all pixel points of the load depth image and the ratio of the element quantity in the pixel point difference characteristic value sequence as an inclined change degree, and recording the product of the inclined characteristic value and the inclined change degree of the load depth image as the ground elastic significance of the corresponding target compacted soil area.
Further, the method for acquiring the significance of the spring soil comprises the following steps:
and carrying out direct proportion normalization on the ratio of the molding compression degree to the ground elasticity significance in the same soil pressing area by adopting a hyperbolic tangent function to obtain the spring significance.
Further, the method for acquiring the pressure change height map comprises the following steps:
obtaining the original depth image and the load depth image which correspond to each soil compacting area and have the same size; and recording the pixel values of the target pixel points in the original depth image and the load depth image in the same soil compaction area as the pixel values of the target pixel points in the pressure change height map, and changing the target pixel points to obtain the pixel value of each pixel point in the pressure change height map.
The invention has the following beneficial effects:
the invention analyzes the soil body according to the elastic deformation characteristic and the inclination characteristic of the spring soil on the basis of the soil body difference corresponding to the spring soil and the normal soil. In order to embody the elastic deformation characteristic of the spring soil, the shaping compression degree is obtained according to the pixel value distribution characteristic in the pressure change height image, the pixel value change characteristic difference between the images before and after rolling in the soil compacting area is introduced on the basis of the shaping compression degree to obtain the shaping flatness, and the shaping deformation degree is obtained according to the shaping flatness and the shaping compression degree, so that the characterization of the shaping deformation degree on the elastic deformation characteristic of the spring soil is more accurate. Furthermore, the method obtains the ground elasticity significance of each soil compaction area according to the element value distribution trend characteristics in the pixel point characteristic value sequence through the inclination characteristics of the rolled spring soil, and characterizes the inclination characteristics of the spring soil through the ground elasticity significance. And further obtaining the spring soil significance according to the plastic deformation degree and the ground elasticity significance of each soil compaction area, wherein the spring soil significance is combined with the elastic deformation characteristic and the inclination characteristic of the soil body, so that the detection accuracy of the spring soil is higher. Furthermore, the method for detecting the spring soil through the image processing has lower cost compared with the method for detecting the spring soil through the neural network in the prior art. In conclusion, the detection cost is reduced while the detection accuracy of the spring soil is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flowchart of a method for detecting spring soil based on computer vision according to an embodiment of the present invention;
fig. 2 is a schematic view of a load depth image direction analysis according to an embodiment of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description will be given to a method for detecting spring soil based on computer vision according to the present invention, with reference to the accompanying drawings and preferred embodiments, and the detailed implementation, structure, features and effects thereof are described below. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following describes a specific scheme of the spring soil detection method based on computer vision in detail with reference to the accompanying drawings.
Referring to fig. 1, a flow chart of a method for detecting spring soil based on computer vision according to an embodiment of the present invention is shown, where the method includes:
step S1: and acquiring an original depth image before load and a load depth image after load of each soil compacting area, and acquiring a pressure change height map according to the difference between the original depth image and the load depth image of each soil compacting area.
The invention aims to perform image processing on images before and after rolling according to the height change characteristic and the inclination characteristic of the soil body after rolling by the soil compactor and further realize the detection of the spring soil according to the image processing result. The precondition for image processing is to obtain the image of the processing target, and considering that the rolled spring soil shows height change or depth change on the physical characteristic, the effect of image processing by adopting the depth camera to collect the soil body image is better than that of an industrial camera reflecting the soil body color, so the depth camera is adopted as the image collecting device.
According to the method for detecting the spring soil, provided by the invention, depth images before and after rolling in the same area need to be shot according to the height change of the soil body before and after rolling, so that the same depth cameras are arranged in front of and behind the soil compactor. In order to ensure the accuracy of the depth image shot by the depth camera and facilitate subsequent analysis, the shooting angle of the depth camera is set to be vertical downward, the horizontal heights of the two depth cameras are consistent, and in addition, the arrangement positions of the two depth cameras in the horizontal direction are required to be ensured on the symmetry axis of the soil compactor. The method comprises the steps of shooting an original depth image corresponding to a target land before a soil compactor rolls the target soil body, shooting a load depth image corresponding to the target land after the soil compactor rolls the target soil body, wherein the original depth image corresponding to the target soil body corresponds to the load depth image in pixel point positions one by one, and the pixel value corresponding to each pixel point in the original depth image and the load depth image is the distance between a depth camera and the soil body surface at the moment of shooting, and the unit is millimeter.
Further, in order to enable the front and rear depth cameras to capture depth images of the same soil compaction area, when the travel distance of the soil compactor coincides with the distance between the front and rear depth cameras, the corresponding front and rear depth cameras capture one time, respectively. In an embodiment of the invention, the travel distance of the compactor is obtained from the travel time and the instantaneous speed of the compactor. And further acquiring an original depth image before the load of each soil compacting area and a load depth image after the load according to the depth camera.
And obtaining an original depth image and a load depth image of each soil compacting area through a depth camera. Considering that the spring soil has the elastic deformation characteristic that the soil body cannot be directly shaped due to the fact that the water content of the spring soil is too high, and the variation difference of the spring soil before and after corresponding rolling is too small compared with that of a normal soil body, the pixel value difference between the original depth image and the load depth image of each soil compaction area can be used as the condition for detecting the spring soil. Specifically, the method comprises the following steps: and obtaining a pressure change height map according to the difference between the original depth image and the load depth image of each soil compacting area. That is, in the same soil compaction area, the pixel value of each pixel point in the pressure change height map is the difference value between the pixel values of the pixel points at the corresponding positions in the original depth image and the load depth image. Preferably, obtaining an original depth image and a load depth image which correspond to each soil compacting area and have the same size; and recording the pixel values of the target pixel points in the original depth image and the load depth image in the same soil compaction area as the pixel values of the target pixel points in the pressure change height map, and changing the target pixel points to obtain the pixel value of each pixel point in the pressure change height map. Expressed on the formula:
wherein the content of the first and second substances,
is a coordinate of
The corresponding pixel value of the pixel point in the compaction change height map;
is a coordinate of
The pixel point of (2) is the corresponding pixel value in the original depth map;
is a coordinate of
The pixel point of (2) is the corresponding pixel value in the depth map after the load.
Step S2: obtaining a concave area and a convex area according to pixel values in the original depth image; and obtaining the molding compression degree of each clay pressing area according to the pixel value distribution characteristics of the pixel points in the concave area and the convex area in the pressure change height diagram.
At this point, an original depth image, a load depth image and a compaction change height map corresponding to each soil compacting area are obtained through the step S1. Because gaps exist among soil particles of a normal soil body, the gaps among the corresponding soil particles can be reduced after the normal soil body is rolled by the road roller, and the physical characteristics show that the height of the soil body rolled by the soil roller is obviously changed relative to the original height; due to the elastic deformation characteristic of the spring soil, the spring soil rolled by the soil compactor can deform towards the shape before rolling, and the physical characteristic shows that the height of the soil body rolled by the soil compactor is not obviously changed relative to the original height. Therefore, the compression characteristics of the soil body are reflected according to the height change of the soil body after rolling, and the characteristics of the spring soil are further represented according to the compression characteristics. However, the height change of the soil body after rolling is not only determined by the soil type of the soil body, but also needs to consider the surface concave-convex degree of the soil body, which is specific:
due to the characteristics of the construction environment, the heights of the soil surface of the construction site are generally uneven. When a normal soil body is rolled by the soil compactor, the convex part of the soil body is obviously reduced after rolling; the concave part is relatively low, the soil with the convex part can fill the concave part of the soil body after rolling, and the corresponding descending degree is relatively small, so that the concave-convex degree of the soil body before rolling needs to be considered when the compression characteristic of the soil body is embodied, namely, different weights are given to the concave part and the convex part on the surface of the soil body for comprehensive evaluation when the compression characteristic of the soil body is calculated.
The method records the compression characteristics of the soil as the plastic compression degree, obtains the concave area and the convex area according to the pixel value in the original depth image, and further obtains the corresponding plastic compression degree of the soil through the pressure change height map. The original depth image is first required to distinguish between the raised and depressed regions of the earthen region. Preferably, a division threshold is obtained by dividing the pixel value of each pixel point in the original depth image by using an OTSU maximum inter-class variance method, the division threshold is a depth threshold, the pixel points corresponding to the pixel values in the original depth image which are greater than the depth threshold are recorded as concave pixel points, the pixel points corresponding to the pixel values in the original depth image which are less than the depth threshold are recorded as convex pixel points, the regions formed by all the concave pixel points are recorded as concave regions, and the regions formed by all the convex pixel points are recorded as convex regions. It should be noted that the variance method between the maximum classes of OTSU is well known in the art, and is not further defined and described herein.
And further obtaining the molding compression degree of each clay pressing area in the pressure change height diagram according to the pixel value distribution characteristics of the pixel points in the concave area and the convex area. Preferably, in the pressure change height map, calculating the pixel value mean value of all pixel points in the concave area and recording the pixel value mean value as the characteristic value of the concave area, calculating the pixel value mean value of all pixel points in the convex area and recording the pixel value mean value as the characteristic value of the convex area, calculating the difference value between the depth threshold value and the characteristic value of the convex area and recording the difference value as the convex difference value, calculating the difference value between the characteristic value of the concave area and the depth threshold value and recording the difference value as the concave difference value, and taking the ratio of the convex difference value to the concave difference value as the weight of the characteristic value of the convex area; and adding the result of multiplying the weight value by the characteristic value of the convex area to the characteristic value of the concave area to obtain the molding compression degree of the corresponding soil compacting area. Expressed on the formula:
wherein, the first and the second end of the pipe are connected with each other,
in order to shape the compression degree of the soil compacting area,
is the depth threshold value corresponding to the soil compacting area,
the coordinates in the depressed area are
The corresponding pixel value of the pixel point in the pressure change height map,
the sum of the pixel values of all pixel points in the depressed area of the soil compacting area in the pressure change height map is obtained,
the coordinates in the raised area of the soil compaction area are
The corresponding pixel value of the pixel point in the pressure change height map,
the sum of the pixel values of all pixel points in the pressure change height map in the raised area of the soil compacting area,
the number of pixel points in the depressed area of the compacted region,
the number of pixel points in the raised area of the soil compaction area.
The molding compression degree calculated by the formula reflects the height change degree of the soil body after the soil pressing area is rolled by the soil pressing machine, when the height change degree is larger, the plastic deformation of the soil body is more obvious, and the significance degree of the corresponding spring soil is lower. In addition, the formula considers the influence of the concave-convex degree of the soil body, the mean value of the pixel values of different concave-convex areas is used as the compression degree of the corresponding area, different weights are respectively given to the compression degree of the convex area and the compression degree of the concave area to further calculate the shaping compression degree, because the height change corresponding to the normal soil body is not greatly different when the soil body corresponding to the concave area is the spring soil, the weight of the compression degree of the concave area is set to be 1, but when the soil body corresponding to the convex area is the spring soil, the height change corresponding to the normal soil body has an obvious area, so the compression degree of the convex area needs to be reducedThe weight is calculated independently, and the weight of the compression degree of the convex area is expressed as the weight in a formula
The difference of the depth threshold value and the corresponding concave-convex area is used as the weight of the spring soil, so that the calculated characterization of the shaping compression degree on the spring soil characteristics can be more accurate by taking the difference of the depth threshold value and the corresponding concave-convex area as the weight of the convex area. And further, giving corresponding weights to the compression degree of the convex area and the compression degree of the concave area of the compacted soil area respectively, and summing to obtain the molding compression degree of the compacted soil area.
And step S3: and obtaining the shaping flatness according to the pixel value variation characteristic difference between the original depth image and the load depth image, and obtaining the shaping deformation according to the shaping compression degree and the shaping flatness of each soil compacting area.
And obtaining the molding compression degrees of all the soil compacting areas through the step S2, wherein the molding compression degrees are obtained according to the difference between the corresponding height change of the soil body corresponding to the spring soil before and after compression and the height change of the normal soil body before and after compression. In addition, after the normal soil body is rolled by the soil compactor, the surface of the soil body is smoother, but the soil body corresponding to the spring soil is not changed greatly before rolling compared with the soil compactor due to the elastic deformation characteristic of the spring soil after being rolled by the soil compactor. Therefore, the analysis can be carried out according to the change of the flatness of the soil compacting area before and after being compacted by the soil compactor. The flatness degree of the soil body is expressed as the change degree of pixel values of the pixel points on the depth image, namely when the unevenness characteristics of the pits on the surface of the soil body are obvious, the pixel value change characteristics of the corresponding pixel points are large; on the contrary, when the land is relatively flat, the pixel value change characteristic of the corresponding pixel point is small.
According to the method, the flatness change of the soil pressing area before and after rolling is represented through the shaping flatness, and the comparison is needed according to the conditions of the soil pressing area before and after rolling, so that the pixel value change characteristics of the original depth image and the load depth image corresponding to the soil pressing area are needed to be obtained firstly, and the pixel value change characteristics are obtained according to the pixel value distribution change in the image. The specific method for acquiring the pixel value change characteristics comprises the following steps: optionally selecting one of the original depth image and the load depth image as a target image; and taking each pixel point of the target image as a central pixel point, acquiring a gray level co-occurrence matrix according to the pixel value of the pixel point in a preset neighborhood range of the central pixel point, calculating the contrast corresponding to the gray level co-occurrence matrix, and recording the accumulated sum of the contrast corresponding to all the pixel points in the target image as a pixel value change characteristic. In the embodiment of the present invention, the predetermined neighborhood range is 5*5 window range. It should be noted that the method for obtaining the gray level co-occurrence matrix and the method for calculating the contrast of the gray level co-occurrence matrix are well known in the prior art, and are not further defined and described herein.
And after the pixel value change characteristics of the images before and after the soil compaction area is compacted by the soil compactor are obtained through calculation, the shaping flatness is obtained according to the pixel value change characteristic difference between the original depth image and the load depth image. Specifically, the method comprises the following steps: optionally selecting one of the original depth image and the load depth image as a target image; and taking each pixel point of the target image as a central pixel point, acquiring a gray level co-occurrence matrix according to the pixel value of the pixel point in a preset neighborhood range of the central pixel point, calculating the contrast corresponding to the gray level co-occurrence matrix, and recording the accumulated sum of the contrast corresponding to all the pixel points in the target image as a pixel value change characteristic. Expressed on the formula:
wherein the content of the first and second substances,
the shaping flatness of the compacted region is shown,
show seatIs marked as
The pixel value of the pixel point in the original depth image changes the characteristic value,
representing coordinates as
The pixel value of the pixel point in the load depth image changes the characteristic value,
a pixel value variation characteristic representing an original depth image,
a pixel value change characteristic representing a depth image of the payload.
The shaping flatness calculated by the formula can reflect the flatness change of the surface of the soil body after the soil compacting region is rolled by the soil compacting machine, when the flatness change is smaller, the soil body corresponding to the soil compacting region is less in change of the compaction degree after being rolled by the soil compacting machine, the corresponding shaping flatness is smaller, the soft plastic state shown by the soil body is more obvious, and the significance degree corresponding to the spring soil is lower. The difference between the pixel value change characteristics of the original depth image and the load depth image of the soil compaction area is used as the flatness change of the soil surface, the change characteristics of the soil body in the soil compaction area after rolling can be clearly shown, the pixel value change characteristics of the load depth image are used as denominators to limit the change characteristics, and the spring soil characteristics of the soil body can be more obvious.
Therefore, the moulding compression degree and the moulding flatness of the soil compacting area are combined to represent the difference between the normal soil body and the soil body corresponding to the spring soil, and the representation value obtained by combining the moulding compression degree and the moulding flatness is recorded as the moulding deformation degree, and the corresponding combination method comprises the following steps: and taking the product of the plastic compression degree and the plastic flatness degree of each soil compacting area as the plastic deformation degree. Expressed on the formula:
wherein the content of the first and second substances,
in order to realize the plastic deformation degree of the soil compacting area,
in order to shape the compression degree of the soil compacting area,
the molding flatness of the soil compacting area is shown. The moulding deformation value reflects the moulding change degree of the soil compacting area before and after being rolled by the soil compacting machine, and is respectively measured from two aspects of the height compression change degree and the surface flatness change degree, when the moulding compression degree and the moulding flatness degree are larger, the corresponding moulding deformation degree is larger, and the spring soil significance degree of the corresponding soil compacting area is smaller.
And step S4: obtaining a pixel point characteristic value sequence in a load depth image according to the movement direction of the soil compactor; and obtaining the ground elasticity significance of each soil compaction area according to the element value distribution trend characteristics in the pixel point characteristic value sequence.
At this point, the plastic deformation degrees of all the compacted regions are obtained through the step S2 and the step S3. Consider that the soil body that the spring soil corresponds suffers destruction because the capillary hole between the great and soil granule of water content for its inside moisture is influenced toward the ability that outwards permeates and distribute, has further reduced the intensity of soil body, shows that elasticity vibration can appear for the soil body itself in physical characteristics. When the soil compactor rolls on soil bodies corresponding to the spring soil, the periphery of a rolling area can be bulged, the whole volume of the corresponding soil body is not compressed, and the front side and the rear side of the soil compactor are tilted upwards and tilted in the direction away from the vehicle. And when the normal soil body is rolled by the soil compactor, the volume can shrink, and the corresponding inclined characteristic can be ignored. Therefore, the invention analyzes the inclination characteristics presented in the soil body rolling process.
When the soil body is rolled by the soil compactor, if the rolled soil body is spring soil, the images before and after rolling all show obvious inclination characteristics, but the inclination characteristics of the soil body can be influenced due to the poor surface flatness degree of the soil body before rolling and the inconsistent heights of different positions. Therefore, in order to reduce the influence of the unevenness of the soil surface on the inclination characteristic, the method analyzes the soil image after rolling, namely, the load depth image is used as an analysis object of the inclination characteristic.
Firstly, a stress characteristic direction is obtained according to the rolling condition of a soil compaction area represented by the load depth image, and a load depth image which is the shortest distance from the rolling time is further analyzed. When the spring soil characteristics of the soil body in the soil compaction area are obvious, the ground heights shown along the stress characteristic direction are sequentially increased on the corresponding load depth map; on the contrary, when the spring soil characteristics of the soil body in the soil compaction area are not obvious, the ground height shown along the stress characteristic direction on the corresponding load depth map has no obvious change. Because the camera position that obtains the load depth image is fixed, consequently can carry out the atress direction analysis with the firming machine direction of motion as the basis, and then obtain the eigenvalue sequence of pixel in the load depth image.
Preferably, in order to embody the characteristics of the spring soil on the load depth map of the soil compacting area, that is, the characteristics of sequentially increasing ground height along the direction of the stress characteristic, the analysis is performed according to the pixel value characteristics of each row of pixel points or each column of pixel points on the load depth map. Because the load depth image represents the information that the soil compaction area is compacted by the soil compaction machine, the direction opposite to the direction of the soil compaction machine is taken as the stress characteristic direction. Referring to fig. 2, which shows a schematic view of analyzing a direction of a load depth image according to an embodiment of the present invention, a default lower boundary of the load depth image in fig. 2 is a side close to the compactor after collection, and an default upper boundary is a side far from the compactor, because a camera position of the load depth image obtained in the embodiment of the present invention is fixed, a motion direction of the compactor may be considered to be perpendicular to a horizontal boundary of the load depth image, that is, a direction in which the upper boundary of the load depth image vertically points to the lower boundary is considered as a motion direction of the compactor, and a direction of a stress characteristic is that the lower boundary vertically points to the upper boundary, the lower boundary is defined as a stress side, and the upper boundary is defined as a stress opposite side.
The convex or concave of the soil surface corresponding to the soil compacting area is very obvious, namely a large pothole or soil pile occurs, the pixel value of the corresponding pixel point is extremely extreme, and the analysis of the pixel value characteristics of each row of pixel points is influenced. In order to better represent the pixel value characteristics of each row of pixel points, the invention selects the pixel value median of each row of pixel points as the pixel point characteristic value corresponding to the pixel point, arranges the pixel point characteristic values in the order of the stress characteristic direction to obtain a pixel point characteristic value sequence, and can analyze the inclination characteristics of the compacted soil area according to the distribution trend characteristics of the pixel point characteristic values in the pixel point characteristic value sequence.
The pixel point characteristic value sequence can represent the change of pixel values corresponding to the pixel points in the load depth image, so that the inclination characteristic of the current compacted soil area after rolling can be analyzed through the distribution trend characteristic of the element values in the pixel point characteristic value sequence, and when the spring soil characteristic of the soil body in the compacted soil area is obvious, the corresponding pixel point characteristic sequence is an obvious increasing sequence with small volatility; when the spring soil characteristics of the soil body in the soil compaction area are not obvious, the corresponding pixel point characteristic sequence is a sequence without obvious change characteristics. Therefore, in order to embody the characteristics of the spring soil, the numerical change of the characteristic value of the inner pixel point of the pixel point characteristic sequence needs to be analyzed, and then the ground elasticity significance of the soil compaction area is obtained.
Preferably, when the soil body corresponding to the spring soil is rolled, the numerical values of the corresponding pixel points are sequentially increased from the stress side to the opposite side of the stress side, and when the weight of the soil compactor is larger, the autocorrelation of the corresponding sequence is larger, so that the characteristic sequence of the pixel points corresponding to the spring soil has long-term memory. Therefore, the autocorrelation and the element of the pixel point feature sequence of the depth image of the corresponding load of the spring soil are consideredThe method comprises the steps of obtaining a hurst index corresponding to a pixel point characteristic sequence by the aid of the increasing characteristics of elements, enabling the hurst index to reflect autocorrelation of the pixel point characteristic sequence, and enabling the corresponding hurst index to be numerically equal to that of the hurst index when the spring soil characteristic corresponding to a soil compaction area is obvious
Within range and close to 1. And obtaining the hurst index by adopting a polymerization variance method for the pixel point characteristic value sequence, and recording the hurst index of the pixel point characteristic value sequence as the autocorrelation characteristic value of the load depth image. It should be noted that the polymerization variance method is well known in the art, and is not further defined and described herein.
In order to enable the incremental characteristic of the elements to be clearer and to reversely map abnormal data in the pixel point characteristic value sequence, the element difference value between every two adjacent elements in the pixel point characteristic value sequence is further calculated, the element difference value is recorded as a pixel point difference characteristic value, the pixel point difference characteristic value is used as an element to form a sequence according to the sequence of the elements in the pixel point characteristic value sequence, and the sequence is recorded as a pixel point difference characteristic value sequence. Under the condition of analyzing the inclined characteristic, the pixel point difference characteristic value sequence can accurately reflect the change process of the pixel point characteristic value.
And calculating the standard deviation among all elements in the pixel point difference characteristic value sequence and recording the standard deviation as the characteristic value difference standard deviation of the load depth image, wherein the characteristic value difference standard deviation can inversely map whether the pixel point characteristic value has regularity in the change process.
Abnormal data detection is carried out on the pixel point difference characteristic value sequence, the number of abnormal values is counted, the abnormal degree of the pixel point characteristic value sequence can be inversely mapped according to the number of the abnormal values, and the representation of the abnormal degree of the pixel point characteristic value sequence can be more accurate by combining standard deviation. In the embodiment of the invention, the adopted abnormal data detection method is an LOF abnormal data detection method. It should be noted that the method for detecting the abnormal data of the LOF is a conventional technique well known to those skilled in the art, and is not further defined or described herein.
And recording the autocorrelation characteristic value, the pixel point difference characteristic value, the characteristic value difference standard deviation and the abnormal value quantity of the load depth image as the element value distribution trend characteristic in the pixel point characteristic value sequence.
And further acquiring the ground elasticity significance of each soil compaction area according to the element value distribution trend characteristics in the pixel point characteristic value sequence of the load depth image. Preferably, a load depth image of the target compacted soil area is obtained, the product of the characteristic value difference standard deviation and the number of the abnormal values of the load depth image is recorded as an abnormal condition characteristic value, the ratio of the autocorrelation characteristic value of the load depth image to the abnormal condition characteristic value plus a preset constant coefficient is calculated and recorded as an inclined characteristic value, the ratio of the sum of the difference characteristic values of all pixel points of the load depth image to the number of elements in the pixel point difference characteristic value sequence is recorded as an inclined change degree, and the product of the inclined characteristic value and the inclined change degree of the load depth image is recorded as the ground elasticity significance of the corresponding target compacted soil area. Expressed on the formula:
wherein, the first and the second end of the pipe are connected with each other,
the ground elasticity significance corresponding to the soil compaction area,
is the corresponding hurst index for the compacted region,
the difference standard deviation of the characteristic values corresponding to the soil compacting area,
the number of abnormal values corresponding to the soil compacting area,
the number of the pixel point difference characteristic values is,
the difference characteristic value of the pixel point is obtained,
representing the accumulated sum of the difference characteristic values of all the pixels in the pixel difference characteristic value sequence,
the serial number of the pixel point difference characteristic value in the pixel point difference characteristic value sequence,
is a constant coefficient. In the embodiment of the invention, the constant coefficient
Set to 1, constant coefficient
The effect of (c) is to make the formula meaningful in order to ensure that the denominator is not zero. When the characteristics of the soil body spring soil are more obvious, the corresponding characteristics
The larger the size of the tube is,
the larger the corresponding ground resilience is, the more significant.
The calculation formula of the ground elasticity significance considers that when the soil body inclination characteristic is obvious, the corresponding situation that the ground height is increased or decreased is obvious. The method takes the mean value of the difference characteristic values of all the pixel points as the inclination change degree, the larger the mean value is, the larger the inclination degree is, and the influence of error conditions on the inclination characteristic calculation can be reduced while the inclination change degree can be clearly reflected by utilizing an averaging method. In addition, the inclination degree of the soil body corresponding to the spring soil after being rolled is related to the weight of the soil compactor, and the self-correlation of the inclination characteristics of the corresponding soil body is related, so that the hurst index is introduced to represent the influence of the self-correlation. Meanwhile, considering that the abnormal condition when the soil body inclines can affect the calculation of the inclination characteristic, the product of the characteristic value difference standard deviation and the abnormal value quantity is used as the characteristic value of the abnormal condition, and the smaller the characteristic value difference standard deviation and the abnormal value quantity is, the more regular the element distribution in the pixel point difference characteristic value sequence is, namely, the more obvious the inclination corresponding to the opposite side is. The hurst index and the ground elasticity significance are in positive correlation, and the abnormal condition characteristic value and the ground elasticity significance are in negative correlation, so that the ratio of the hurst index to the abnormal condition characteristic value is further used as an inclination characteristic value, the inclination change degree is limited through the inclination characteristic value, the autocorrelation and the abnormal condition of the soil body inclination characteristic are considered, and the inclination characteristic expressed by the ground elasticity significance can be more accurate. The ground elasticity significance calculated by the calculation formula of the ground elasticity significance can accurately reflect the inclination characteristic significance degree of the soil body inclining towards the direction of the soil compactor after the soil body corresponding to the spring soil is rolled by the soil compactor, and when the inclination characteristic significance degree is larger, the ground elasticity significance of the corresponding soil body is larger, namely the possibility that the spring soil appears in the corresponding soil body is higher.
Step S5: and obtaining the spring soil significance of each compacted soil region according to the plastic deformation degree and the ground elasticity significance of each compacted soil region, and completing spring soil detection according to the spring soil significance.
And obtaining the plastic deformation degree and the ground elasticity significance degree corresponding to all the soil compacting areas through the step S2, the step S3 and the step S4, and further obtaining the spring soil significance degree of each soil compacting area according to the plastic deformation degree and the ground elasticity significance of each soil compacting area. The method normalizes the ratio of the ground elasticity significance to the plastic deformation of each soil compaction area to obtain the corresponding spring soil significance. Preferably, according to the method, a hyperbolic tangent function is adopted to normalize the ratio of the plastic compression degree to the ground elasticity significance in the same soil pressing area in a direct proportion manner to obtain the spring significance. It should be noted that the hyperbolic tangent function is well known in the art, and is not further defined and described herein. The calculation process of the significance of the spring soil is represented by a formula as follows:
wherein the content of the first and second substances,
the spring soil prominence of the soil compaction area,
in order to account for the ground elastic significance of the compacted region,
in order to realize the plastic deformation degree of the soil compacting area,
representing a hyperbolic tangent function normalization. When the ground elasticity significance of the soil compaction area is larger, the plastic deformation degree is smaller, and the corresponding spring soil significance degree is larger. The spring soil significance is comprehensively evaluated by the ground elasticity significance and the plastic deformation degree of the soil compaction area to obtain a significant characteristic value, and when the spring soil significance is higher, the possibility that the spring soil appears on the soil body corresponding to the soil compaction area is more obvious.
When the spring soil significance of the soil compaction area is smaller than a preset spring soil threshold value, the soil body corresponding to the current soil compaction area is considered to be a normal soil body, and the normal soil body is not processed; when the spring soil significance of the soil compacting area is greater than or equal to the preset spring soil threshold, the soil body of the current soil compacting area is considered to have spring soil and influence the construction process, and the soil of the soil body needs to be processed in time. In the embodiment of the invention, the preset spring soil threshold value is 0.8.
Thus, the present invention has been completed.
In summary, according to the invention, the original depth image before the soil compaction region is compacted by the soil compactor and the load depth image after compaction are obtained through the depth camera, the pressure change height image is obtained according to the difference between the original depth image and the load depth image, the shaping compression degree is obtained according to the pixel value distribution characteristics in the original depth image and the pressure change height image, the shaping flatness is obtained according to the pixel value change characteristics in the original depth image and the depth image after load, the shaping deformation degree is obtained according to the shaping compression degree and the shaping flatness, the ground elasticity saliency is obtained according to the load depth image, the spring soil saliency is obtained according to the shaping deformation degree and the ground elasticity saliency together, and the spring soil detection is completed according to the spring soil saliency. The invention improves the detection accuracy of the spring soil and reduces the detection cost at the same time.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. The processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
All the embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from other embodiments.
Claims (10)
1. A method for detecting spring soil based on computer vision is characterized by comprising the following steps:
acquiring an original depth image before load and a load depth image after load of each soil compacting area, and acquiring a pressure change height map according to the difference between the original depth image and the load depth image of each soil compacting area;
obtaining a concave area and a convex area according to the pixel value in the original depth image; obtaining the molding compression degree of each soil compacting region in the pressure change height map according to the pixel value distribution characteristics of the pixel points in the concave region and the convex region;
obtaining a shaping flatness according to the pixel value variation characteristic difference between the original depth image and the load depth image, and obtaining a shaping deformation degree according to the shaping compression degree and the shaping flatness degree of each soil compacting area;
obtaining a pixel point characteristic value sequence in the load depth image according to the movement direction of the soil compactor; acquiring the ground elasticity significance of each soil compaction area according to the element value distribution trend characteristics in the pixel point characteristic value sequence;
and obtaining the spring soil significance of each compacted soil area according to the plastic deformation degree and the ground elasticity significance of each compacted soil area, and completing spring soil detection according to the spring soil significance.
2. The method for detecting the spring soil based on the computer vision is characterized in that the method for acquiring the concave area and the convex area of the pressure change height map comprises the following steps:
and dividing the pixel values of all the pixel points in the original depth image by adopting an OTSU maximum inter-class variance method to obtain a division threshold, wherein the division threshold is a depth threshold, the pixel points corresponding to the pixel values in the original depth image which are greater than the depth threshold are recorded as concave pixel points, the pixel points corresponding to the pixel values in the original depth image which are less than the depth threshold are recorded as convex pixel points, the regions formed by all the concave pixel points are recorded as concave regions, and the regions formed by all the convex pixel points are recorded as convex regions.
3. The method for detecting the spring soil based on the computer vision as claimed in claim 2, wherein the method for obtaining the molding compression degree comprises:
in the pressure change height map, calculating the pixel value mean value of all pixel points in the concave area and recording the pixel value mean value as a concave area characteristic value, calculating the pixel value mean value of all pixel points in the convex area and recording the pixel value mean value as a convex area characteristic value, calculating the difference value between the depth threshold value and the convex area characteristic value and recording the difference value as a convex difference value, calculating the difference value between the concave area characteristic value and the depth threshold value and recording the difference value as a concave difference value, and taking the ratio of the convex difference value to the concave difference value as the weight of the convex area characteristic value; and adding the result of multiplying the weight by the characteristic value of the convex area to the characteristic value of the concave area to obtain the plastic compression degree of the corresponding soil compacting area.
4. The method for detecting the spring soil based on the computer vision as claimed in claim 1, wherein the method for acquiring the pixel value variation characteristic comprises:
optionally selecting one of the original depth image and the load depth image as a target image;
taking each pixel point of the target image as a central pixel point, acquiring a gray level co-occurrence matrix according to the pixel value of the pixel point in a preset neighborhood range of the central pixel point, calculating the contrast corresponding to the gray level co-occurrence matrix, and recording the accumulated sum of the contrast corresponding to all the pixel points in the target image as a pixel value change characteristic.
5. The method for detecting the spring soil based on the computer vision as claimed in claim 1, wherein the method for obtaining the shaping flatness comprises:
calculating the difference between the pressure change height map and the pixel value change characteristics of the load depth image in the same soil compaction area, recording the difference as the flatness difference, and recording the ratio of the flatness difference to the pixel value change characteristics of the load depth image as the shaping flatness of the corresponding soil compaction area.
6. The method for detecting the spring soil based on the computer vision as claimed in claim 1, wherein the method for obtaining the pixel point feature value sequence comprises:
in the load depth image, taking the direction opposite to the movement direction of the soil compactor as the stress characteristic direction;
and counting the pixel value median of each row of pixels in the load depth image, recording the pixel value median as a pixel characteristic value, and arranging the pixel characteristic values in a stress characteristic direction as a sequence to obtain a pixel characteristic value sequence.
7. The method for detecting the spring soil based on the computer vision as claimed in claim 6, wherein the method for obtaining the element value distribution trend characteristics in the pixel point characteristic value sequence comprises:
obtaining a hurst index by adopting a polymerization variance method for the pixel point characteristic value sequence, and recording the hurst index of the pixel point characteristic value sequence as an autocorrelation characteristic value of the load depth image;
calculating element difference values between every two adjacent elements in the pixel point characteristic value sequence, recording the element difference values as pixel point difference characteristic values, forming a sequence by using the pixel point difference characteristic values as elements according to the sequence of the elements in the pixel point characteristic value sequence, and recording the sequence as the pixel point difference characteristic value sequence;
calculating the standard deviation among various elements in the pixel point difference characteristic value sequence, recording the standard deviation as the characteristic value difference standard deviation of the load depth image, carrying out abnormal data detection on the pixel point difference characteristic value sequence and counting the number of abnormal values;
and recording the autocorrelation characteristic value, the pixel point difference characteristic value, the characteristic value difference standard deviation and the abnormal value quantity of the load depth image as the distribution trend characteristic.
8. The computer vision-based spring soil detection method according to claim 7, wherein the method for acquiring the ground elasticity significance comprises the following steps:
acquiring the load depth image of the target compacted soil area, recording the product of the characteristic value difference standard deviation and the abnormal value quantity of the load depth image as an abnormal condition characteristic value, calculating the ratio of the autocorrelation characteristic value of the load depth image to the abnormal condition characteristic value plus a preset constant coefficient, recording the ratio as an inclined characteristic value, recording the sum of the difference characteristic values of all pixel points of the load depth image and the ratio of the element quantity in the pixel point difference characteristic value sequence as an inclined change degree, and recording the product of the inclined characteristic value and the inclined change degree of the load depth image as the ground elastic significance of the corresponding target compacted soil area.
9. The method for detecting the spring soil based on the computer vision is characterized in that the method for acquiring the significance of the spring soil comprises the following steps:
and carrying out direct proportion normalization on the ratio of the moulding compression degree to the ground elasticity significance in the same soil pressing area by adopting a hyperbolic tangent function to obtain the spring significance.
10. The method for detecting the spring soil based on the computer vision is characterized in that the method for acquiring the pressure change height map comprises the following steps:
obtaining the original depth image and the load depth image which correspond to each soil compacting area and have the same size; and recording the pixel values of the target pixel points in the original depth image and the load depth image in the same soil compaction area as the pixel values of the target pixel points in the pressure change height map, and changing the target pixel points to obtain the pixel value of each pixel point in the pressure change height map.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310139076.1A CN115830029B (en) | 2023-02-21 | 2023-02-21 | Spring soil detection method based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310139076.1A CN115830029B (en) | 2023-02-21 | 2023-02-21 | Spring soil detection method based on computer vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115830029A true CN115830029A (en) | 2023-03-21 |
| CN115830029B CN115830029B (en) | 2023-04-28 |
Family
ID=85521947
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310139076.1A Active CN115830029B (en) | 2023-02-21 | 2023-02-21 | Spring soil detection method based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115830029B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103033456A (en) * | 2012-12-13 | 2013-04-10 | 北京农业信息技术研究中心 | Soil porosity detection method based on SFS (Shape from Shading) algorithm |
| CN104103082A (en) * | 2014-06-06 | 2014-10-15 | 华南理工大学 | Image saliency detection method based on region description and priori knowledge |
| US20150117784A1 (en) * | 2013-10-24 | 2015-04-30 | Adobe Systems Incorporated | Image foreground detection |
| CN106872472A (en) * | 2017-01-17 | 2017-06-20 | 中交武汉港湾工程设计研究院有限公司 | Surface Quality of Concrete method of determination and evaluation |
| US20170351941A1 (en) * | 2016-06-03 | 2017-12-07 | Miovision Technologies Incorporated | System and Method for Performing Saliency Detection Using Deep Active Contours |
| CN111145161A (en) * | 2019-12-28 | 2020-05-12 | 北京工业大学 | Method for processing and identifying pavement crack digital image |
| CN111932537A (en) * | 2020-10-09 | 2020-11-13 | 腾讯科技（深圳）有限公司 | Object deformation detection method and device, computer equipment and storage medium |
| CN114782419A (en) * | 2022-06-17 | 2022-07-22 | 山东水利建设集团有限公司 | Water conservancy construction gradient detection method |
| CN114972906A (en) * | 2022-05-05 | 2022-08-30 | 同济大学 | Soil quality type identification method for excavation surface of soil pressure balance shield |
-
2023
- 2023-02-21 CN CN202310139076.1A patent/CN115830029B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103033456A (en) * | 2012-12-13 | 2013-04-10 | 北京农业信息技术研究中心 | Soil porosity detection method based on SFS (Shape from Shading) algorithm |
| US20150117784A1 (en) * | 2013-10-24 | 2015-04-30 | Adobe Systems Incorporated | Image foreground detection |
| CN104103082A (en) * | 2014-06-06 | 2014-10-15 | 华南理工大学 | Image saliency detection method based on region description and priori knowledge |
| US20170351941A1 (en) * | 2016-06-03 | 2017-12-07 | Miovision Technologies Incorporated | System and Method for Performing Saliency Detection Using Deep Active Contours |
| CN106872472A (en) * | 2017-01-17 | 2017-06-20 | 中交武汉港湾工程设计研究院有限公司 | Surface Quality of Concrete method of determination and evaluation |
| CN111145161A (en) * | 2019-12-28 | 2020-05-12 | 北京工业大学 | Method for processing and identifying pavement crack digital image |
| CN111932537A (en) * | 2020-10-09 | 2020-11-13 | 腾讯科技（深圳）有限公司 | Object deformation detection method and device, computer equipment and storage medium |
| CN114972906A (en) * | 2022-05-05 | 2022-08-30 | 同济大学 | Soil quality type identification method for excavation surface of soil pressure balance shield |
| CN114782419A (en) * | 2022-06-17 | 2022-07-22 | 山东水利建设集团有限公司 | Water conservancy construction gradient detection method |
Non-Patent Citations (3)
| Title |
|---|
| ZAMPELA PITTAKI CHRYSODONTA ET AL.: "Rapid estimation of a soil – water retention curve using visible – near infrared spectroscopy" * |
| 李彦飞: "路基"弹簧土"的工程处理" * |
| 魏锦山: "基于深度学习与近红外光谱的土壤分类方法研究" * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115830029B (en) | 2023-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP3426060B2 (en) | Face image processing device | |
| CN105510195A (en) | On-line detection method for particle size and shape of stacked aggregate | |
| CN110611770B (en) | Method and system for judging whether line frequency of linear array camera is matched with object motion speed | |
| CN110533686B (en) | Method and system for judging whether line frequency of linear array camera is matched with object motion speed | |
| CN107328502B (en) | Anchor rod tray load visualization digital imaging method | |
| CN111274939B (en) | Automatic extraction method for road pavement pothole damage based on monocular camera | |
| CN116758059A (en) | Visual nondestructive testing method for roadbed and pavement | |
| CN110346151B (en) | Apparatus and computer-implemented method for determining the mass of a tire | |
| CN113605188B (en) | Pavement structure testing method | |
| CN106503643A (en) | Tumble detection method for human body | |
| CN107341790A (en) | A kind of image processing method of environment cleanliness detection | |
| CN107784646A (en) | A kind of road self-adapting detecting method to gather materials | |
| CN116580032B (en) | Quality monitoring method for road construction | |
| CN107340298A (en) | Balance car system monitoring method based on camera pavement detection | |
| CN115830029A (en) | Spring soil detection method based on computer vision | |
| CN116399302B (en) | Method for monitoring dynamic compaction settlement in real time based on binocular vision and neural network model | |
| CN101719271A (en) | Video shot boundary detection method based on mixed projection function and support vector machine | |
| CN111612734B (en) | Background clutter characterization method based on image structure complexity | |
| CN116363617A (en) | Ultra-small curvature lane line detection method | |
| CN107977608A (en) | A kind of method applied to the extraction of highway video image road area | |
| CN105550514A (en) | Dual-path integration based speed model establishment method and system | |
| CN105300289A (en) | Vision measurement method for wheel settlement amount of planet vehicle in complex terrain | |
| TWI750950B (en) | Camera angle detection method and related surveillance apparatus | |
| CN112284287B (en) | Stereoscopic vision three-dimensional displacement measurement method based on structural surface gray scale characteristics | |
| CN115115498A (en) | Excavator operation material judgment device and method based on visual identification |
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 |