CN112767471B - Tire ground contact area measuring method and device based on image feature extraction - Google Patents

Tire ground contact area measuring method and device based on image feature extraction Download PDF

Info

Publication number
CN112767471B
CN112767471B CN202110003922.8A CN202110003922A CN112767471B CN 112767471 B CN112767471 B CN 112767471B CN 202110003922 A CN202110003922 A CN 202110003922A CN 112767471 B CN112767471 B CN 112767471B
Authority
CN
China
Prior art keywords
feature
image
tire
saliency map
map
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
Application number
CN202110003922.8A
Other languages
Chinese (zh)
Other versions
CN112767471A (en
Inventor
孔烜
易金鑫
邓露
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hunan University
Original Assignee
Hunan University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Hunan University filed Critical Hunan University
Priority to CN202110003922.8A priority Critical patent/CN112767471B/en
Publication of CN112767471A publication Critical patent/CN112767471A/en
Application granted granted Critical
Publication of CN112767471B publication Critical patent/CN112767471B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds

Abstract

The application discloses a tire ground contact area measuring method, device, equipment and computer readable storage medium based on image feature extraction, wherein the method comprises the following steps: acquiring an image of a target tire, extracting a feature map from the image and generating a feature saliency map; calculating the optimal weight value of each characteristic saliency map according to the target tire in the image, and carrying out normalization processing on each optimal weight value; according to the feature saliency map and the corresponding optimal weight value after normalization processing, carrying out threshold segmentation on the saliency map to obtain a tire region image; and acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length. According to the technical scheme, the ground contact area of the tire is automatically and contactlessly measured by acquiring the image of the target tire and processing the image of the target tire, so that the efficiency, accuracy and safety of the measurement of the ground contact area of the tire are improved.

Description

Tire ground contact area measuring method and device based on image feature extraction
Technical Field
The present disclosure relates to the field of vehicle measurement, and more particularly, to a tire ground contact area measurement method, apparatus, device, and computer readable storage medium based on image feature extraction.
Background
The standard vehicle single-wheel pressure transmission surface equivalent circle diameter is 21.30+/-0.5 cm specified in JTG E60-2008 of the highway subgrade road surface field test rule, and the single-wheel pressure transmission surface equivalent circle diameter needs to be converted through the tire ground area.
Currently, the tire footprint is often determined by: the rear axle of the automobile is jacked up by a jack on a smooth hard road surface, a piece of new duplicating paper and a piece of square paper are paved below the tire, the jack is gently dropped down, then the jack is jacked up, namely, tire marks are printed on the square paper, and the grounding area of the tire is measured by a integrating meter or a method for counting squares. The above measurement process requires a lot of manpower and time, and when jack-up jack and uninstallation jack, the operating personnel need to squat at the vehicle bottom and when placing and taking out duplicating paper and square paper, the operating personnel need be close to the wheel, these all have increased the potential safety hazard to operating personnel, and simultaneously because the jack uninstallation speed is faster, the vehicle produces impact force easily, makes the measurement tire area bigger.
In summary, how to improve the efficiency, accuracy and safety of the tire ground contact area measurement is a technical problem to be solved by those skilled in the art.
Disclosure of Invention
In view of the foregoing, it is an object of the present application to provide a tire ground contact area measurement method, apparatus, device and computer readable storage medium based on image feature extraction for improving the efficiency, accuracy and safety of tire ground contact area measurement.
In order to achieve the above object, the present application provides the following technical solutions:
a tire ground contact area measuring method based on image feature extraction comprises the following steps:
acquiring an image of a target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map;
calculating the optimal weight value of each characteristic saliency map according to the target tire in the image, and carrying out normalization processing on each optimal weight value;
obtaining a salient map of the image according to each characteristic salient map and the corresponding normalized optimal weight value, and performing threshold segmentation on the salient map of the image to obtain a tire area image;
and acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length.
Preferably, acquiring the contact length of the target tire with the ground according to the tire area image includes:
identifying a specification mark of the target tire, the number of pixels corresponding to the rim diameter and the number of grounding pixels of the target tire from the tire area image, and acquiring the rim diameter from the specification mark;
calculating a scale factor of the image by using the rim diameter and the number of pixels corresponding to the rim diameter;
and calculating the contact length of the target tire and the ground by using the number of the grounding pixels of the target tire and the scale factor of the image.
Preferably, processing each of the feature graphs to generate a corresponding feature saliency map includes:
and processing each characteristic map through a Gaussian pyramid and a central peripheral difference operator to generate a characteristic saliency map corresponding to each characteristic map.
Preferably, extracting a plurality of feature maps from the image includes:
and extracting a color feature map, a texture feature map, a brightness feature map and a shape feature map from the image.
Preferably, obtaining the saliency map of the image according to each feature saliency map and the corresponding normalized optimal weight value includes:
Obtaining a saliency map of the image using i=g (C) c+g (T) t+g (B) b+g (S) S;
wherein I is a saliency map of the image, C is a color feature saliency map, g (C) is an optimal weight value after normalization processing corresponding to the color feature saliency map, T is a texture feature saliency map, g (T) is an optimal weight value after normalization processing corresponding to the texture feature saliency map, B is a brightness feature saliency map, g (B) is an optimal weight value after normalization processing corresponding to the brightness feature saliency map, S is a shape feature saliency map, and g (S) is an optimal weight value after normalization processing corresponding to the shape feature saliency map.
Preferably, calculating an optimal weight value of each feature saliency map according to the target tire in the image includes:
and calculating the optimal weight value of each characteristic saliency map by using a genetic algorithm according to the target tire in the image.
Preferably, calculating an optimal weight value of each feature saliency map according to a target tire in the image by using a genetic algorithm includes:
randomly selecting a first preset number of feature weights as a first generation of individuals for each feature saliency map;
calculating a characteristic ROI image under each individual according to each individual and the characteristic saliency map, and comparing the characteristic ROI image under each individual with the image in a similarity mode; the similarity of the region of interest is obtained by multiplying the point-to-point line similarity and the position similarity;
Selecting two feature weights corresponding to the feature ROI images with the highest image similarity as target feature weights, and judging whether the two target feature weights meet a set condition or not;
if not, randomly selecting a second preset number of feature weights between the two target feature weights, taking the two target feature weights and the second preset number of feature weights as next generation individuals, and executing the step of calculating a feature ROI image under each individual according to each individual and the feature saliency map; wherein the first preset number differs from the second preset number by two;
if yes, calculating the average value of the two target feature weights, and taking the average value of the two target feature weights as the optimal weight value of the feature saliency map.
A tire ground contact area measurement apparatus based on image feature extraction, comprising:
the acquisition module is used for acquiring an image of the target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map;
the first calculation module is used for calculating the optimal weight value of each characteristic significant figure according to the target tire in the image and carrying out normalization processing on each optimal weight value;
The tire region image obtaining module is used for obtaining a significant image of the image according to each characteristic significant image and the corresponding normalized optimal weight value, and carrying out threshold segmentation on the significant image of the image to obtain a tire region image;
and the second calculation module is used for acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length.
A tire ground contact area measurement apparatus based on image feature extraction, comprising:
a memory for storing a computer program;
a processor for implementing the steps of the tire footprint measurement method based on image feature extraction as described in any one of the above when executing the computer program.
A computer readable storage medium having stored therein a computer program which, when executed by a processor, implements the steps of the tire footprint measurement method based on image feature extraction as set forth in any one of the preceding claims.
The application provides a tire ground contact area measuring method, device and equipment based on image feature extraction and a computer readable storage medium, wherein the method comprises the following steps: acquiring an image of a target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map; calculating the optimal weight value of each characteristic saliency map according to the target tire in the image, and carrying out normalization processing on each optimal weight value; obtaining a salient image of the image according to each characteristic salient image and the corresponding normalized optimal weight value, and carrying out threshold segmentation on the salient image of the image to obtain a tire area image; and acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length.
According to the technical scheme, the multiple feature images are extracted from the image of the target tire and processed to generate the corresponding feature saliency map, the optimal weight value of each feature saliency map is calculated according to the target tire in the image of the target tire, normalization processing is carried out on each optimal weight value, the saliency map of the image of the target tire is calculated according to each feature saliency map and the corresponding normalized optimal weight value, threshold segmentation is carried out on the saliency map to obtain the tire area image, then the width of the target tire and the contact length of the target tire and the ground are obtained according to the tire area image, and the ground contact area of the tire is automatically calculated in a non-contact mode according to the width of the target tire and the contact length of the target tire and the ground without artificial participation, so that the efficiency and the accuracy of the measurement of the tire ground contact area are improved, the potential safety hazards existing in the measurement process are reduced, and the safety of the measurement is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present application, and that other drawings may be obtained according to the provided drawings without inventive effort to a person skilled in the art.
Fig. 1 is a flowchart of a tire ground contact area measurement method based on image feature extraction according to an embodiment of the present application;
fig. 2 is a schematic diagram of pixel local maximum value determination provided in an embodiment of the present application;
fig. 3 is a schematic structural diagram of a tire ground contact area measurement device based on image feature extraction according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a tire ground contact area measurement apparatus based on image feature extraction according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
Referring to fig. 1, which is a flowchart illustrating a tire ground area measurement method based on image feature extraction provided in an embodiment of the present application, the tire ground area measurement method based on image feature extraction provided in an embodiment of the present application may include:
S11: and acquiring an image of the target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map.
Considering the problem that the existing mode of stamping on the copying paper and the check paper by means of a jack and measuring the tire contact area according to the stamping is low in efficiency, accuracy and safety, the application provides a tire contact area measuring method based on image feature extraction, which is used for improving the efficiency, accuracy and safety of tire contact area measurement.
Specifically, when the ground contact area measurement is required for the target tire, the image of the target tire may be acquired first, and a plurality of feature maps may be extracted from the image of the target tire and may be processed to generate feature saliency maps corresponding to the feature maps, respectively, so that the features of the target tire may be more salient, thereby facilitating accurate acquisition of the tire region image of interest according to the plurality of feature saliency maps.
It should be noted that, in the present application, the camera disposed on the side of the vehicle driving road may be specifically used to obtain the image of the target tire, where the lens of the camera is perpendicular to the driving direction of the vehicle, so as to capture the image of the front surface of the target tire, so as to facilitate improvement of the quality of the image of the target tire, and further facilitate improvement of the accuracy of the measurement of the ground contact area of the target tire.
S12: and calculating the optimal weight value of each characteristic saliency map according to the target tire in the image, and carrying out normalization processing on each optimal weight value.
After step S11 is performed, calculating an optimal weight value of each feature saliency map according to the target tire in the image of the target tire, and normalizing the optimal weight value of each feature saliency map, where the normalization process mentioned herein is to divide the sum of the optimal weight value of each feature saliency map and the optimal weight value of all feature saliency maps, so as to obtain the normalized optimal weight value corresponding to each feature saliency map.
The importance of each feature saliency map can be represented through the optimal weight value, and the tire region image of interest can be conveniently obtained according to the normalized optimal weight value corresponding to each feature saliency map and the corresponding feature saliency map.
S13: according to the feature saliency maps and the corresponding normalized optimal weight values, obtaining a saliency map of the image, and performing threshold segmentation on the saliency map of the image to obtain a tire area image.
After obtaining the optimal weight value after normalization processing corresponding to each feature saliency map, the feature saliency maps and the optimal weight value after normalization processing corresponding to each feature saliency map can be linearly combined to obtain a saliency map of an image of a target tire, and then the saliency map of the image of the target tire can be subjected to threshold segmentation to obtain a tire region image, wherein the threshold value during threshold segmentation can be empirically set according to the situation of the target tire, and the tire region image only comprises the target tire and does not comprise a background, so that the background is prevented from influencing the measurement of the tire ground contact area, and the accuracy of the measurement of the tire ground contact area is improved conveniently.
S14: and acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length.
After the tire area image is extracted, the specification mark of the target tire may be identified from the tire area image, and the width of the target tire may be obtained from the specification mark of the target tire, while the contact length of the target tire with the ground is obtained from the tire area image, and then the width of the target tire is multiplied by the contact length of the target tire with the ground to obtain the ground contact area of the target tire.
According to the process, the ground contact area of the target tire can be measured in a non-contact manner by acquiring and processing the image of the target tire, the operation is simple and convenient without human participation, and a large amount of manpower and time are not required to be consumed, so that the speed of measuring the ground contact area of the tire is relatively high, the efficiency is relatively high, the accuracy and the precision are relatively high, unsafe factors such as squatting at the bottom of a vehicle are avoided, and the safety of measuring the ground contact area of the tire can be improved.
According to the technical scheme, the multiple feature images are extracted from the image of the target tire and processed to generate the corresponding feature saliency map, the optimal weight value of each feature saliency map is calculated according to the target tire in the image of the target tire, normalization processing is carried out on each optimal weight value, the saliency map of the image of the target tire is calculated according to each feature saliency map and the corresponding normalized optimal weight value, threshold segmentation is carried out on the saliency map to obtain the tire area image, then the width of the target tire and the contact length of the target tire and the ground are obtained according to the tire area image, and the ground contact area of the tire is automatically calculated in a non-contact mode according to the width of the target tire and the contact length of the target tire and the ground without artificial participation, so that the efficiency and the accuracy of the measurement of the tire ground contact area are improved, the potential safety hazards existing in the measurement process are reduced, and the safety of the measurement is improved.
The method for measuring the tire ground contact area based on image feature extraction provided by the embodiment of the application, according to the contact length of the tire region image acquisition target tire and the ground, may include:
Identifying a specification mark of the target tire, the number of pixels corresponding to the rim diameter and the number of grounding pixels of the target tire from the tire area image, and acquiring the rim diameter from the specification mark;
calculating the scale factor of the image by using the rim diameter and the number of pixels corresponding to the rim diameter;
and calculating the contact length of the target tire and the ground by using the number of the ground contact pixels of the target tire and the scale factor of the image.
In the present application, the contact length of the target tire with the ground can be obtained from the tire area image specifically by: the method comprises the steps of identifying the number of pixels corresponding to the rim diameter and the number of grounding pixels of a target tire from a tire area image, identifying the specification mark of the target tire from the tire area image, obtaining the rim diameter from the specification mark of the target, dividing the rim diameter by the number of pixels corresponding to the rim diameter to obtain a scale factor of the image, wherein the scale factor represents the actual length corresponding to one pixel point in the image of the target tire, and multiplying the number of grounding pixels of the target tire by the scale factor of the image to calculate the contact length of the target tire and the ground.
Because the rim is a rigid body, when the tire is deformed, the rim is hardly deformed, so that the accuracy of calculating the scale factor can be improved when the diameter of the rim is used for calculating the scale factor, the accuracy of calculating the contact length of the target tire and the ground can be improved conveniently, and the accuracy of measuring the ground contact area of the tire can be improved conveniently.
The method for measuring the tire ground contact area based on image feature extraction provided by the embodiment of the application processes each feature map to generate a corresponding feature saliency map, and may include:
and processing each feature map through a Gaussian pyramid and a central peripheral difference operator to generate a feature saliency map corresponding to each feature map.
When each feature map is processed to generate a corresponding feature saliency map, each feature map can be processed through a Gaussian pyramid and a central peripheral difference operator respectively, so that feature saliency maps corresponding to each feature map are generated respectively, the features of a target tire are more obvious, the saliency of a tire area image is improved conveniently, and the accuracy of the measurement of the tire ground contact area is improved conveniently.
The specific process for generating the feature saliency map corresponding to the feature map is as follows: for an image I of width w and height h, the gaussian pyramid is composed of a gaussian image I (σ) obtained by reducing the resolution thereof, where σ= {0, 1. Each stage of image I (sigma) is obtained by carrying out Gaussian smoothing processing and interlaced column sampling on an original image I (0) step by step downwards. A better scale c is selected from the gaussian pyramid I (σ) of the image I to represent "center", and a coarser scale s is selected as "periphery", so that the difference between the corresponding positions of the two scales is called the center-periphery difference. The greater the difference between the center and the periphery, the more pronounced.
The method for measuring the tire ground contact area based on image feature extraction provided by the embodiment of the application, extracting a plurality of feature images from an image, may include:
and extracting a color feature map, a texture feature map, a brightness feature map and a shape feature map from the image.
When extracting a plurality of feature maps from an image, a color feature map, a texture feature map, a brightness feature map and a shape feature map can be specifically extracted from the image, wherein the color feature is a global feature, describes the surface property of a scene corresponding to the image or the image region, the texture feature is a global feature, also describes the surface property of the scene corresponding to the image or the image region, the brightness feature refers to the brightness degree of each part of the image, the shape feature has two types of representation methods, one type is a contour feature, the other type is a region feature, the contour feature of the image is mainly aimed at the outer boundary of an object, and the shape feature in the application is mainly referred to as a contour feature.
Wherein, 1) the process of extracting the color feature map is as follows:
converting the picture into HIS (H-hue, S-saturation, I-intensity or brightness) mode, and then extracting a color feature map of the sample picture by the following formula:
Wherein f c (x, y) is the gray value of the obtained color feature map at (x, y), and saturation (x, y) and brightness (x, y) respectively represent the saturation value and brightness value of the pixel at (x, y) ave And bright ness ave Representing the saturation average and the brightness average of the whole image, s c And b c Are constant and can be taken to be 0.5.
It should be noted that, because the bright and purer color area is more attractive than the dark and complex and diversified color area, the above characteristics can be represented by the above formula, and the influence of brightness and saturation on the color characteristics is comprehensively considered by the formula.
2) The process for extracting the brightness characteristic map comprises the following steps:
the image of the target tire is decomposed into three monochromatic images R (red), G (green) and B (blue), pixels of each monochromatic image have different gray values, and the brightness value of each pixel of the image is generated through the formula I=0.2989R+0.5871G+0.1140B, so that a brightness characteristic diagram of the sample image can be obtained.
3) The process for extracting the texture feature map comprises the following steps:
the two-dimensional filter is adopted to extract texture characteristics of the image, and different parameters are set to form different two-dimensional Gabor functions, which essentially are to convolve the image. The corresponding impulse response function of a two-dimensional Gabor filter is generally shown as follows:
Wherein k is x ,k y Representing the relative frequency components along the x-axis and y-axis directions, respectively, delta representing the standard deviation of the gaussian function,represents the relative frequency of the radial direction, +.>The direction of the Gabor filter is expressed, different filtering templates can be obtained by taking different theta values, and different texture features can be obtained by selecting a group of filters with different dominant frequencies for feature extraction.
The wavelet transform can multiscale analyze the signal, and the combination of the multiscale of the Gabor transform and the multiscale of the wavelet transform produces a Gabor wavelet transform that can multiscale, multidirectional extract the texture features of the image. It is essentially based on Gabor transformation, followed by wavelet transformation. The Gabor function is used as a basis function, combined with wavelet transformation in a non-standard orthogonal basis mode, and convolved with an image, and localized frequency information can be obtained through expansion of the basis function. Then, the Gabor wavelet transformation is utilized to extract texture features of the target tire in 5 dimensions and 8 directions, and weights such as 40 texture maps are combined to form a texture feature map.
4) The process for extracting the shape feature map comprises the following steps:
extracting shape characteristics of a sample picture by using a Canny operator, wherein the method comprises the following four steps:
41 Noise reduction): gaussian filtering is used in the Canny operator to improve edge detection performance. The original image data is convolved with a gaussian smoothing template to eliminate noise, but the resulting image is slightly blurred compared to the original image. In this way, the noise of a single pixel becomes hardly affected after processing.
42 Calculating gradient direction and magnitude: the gradient of the gray value of the image is approximated by utilizing the first order difference to obtain partial derivative matrixes in the x and y directions, a pair of convolution arrays are distributed to act on the corresponding horizontal and vertical directions, and the following Sobel template is used:
wherein G is x Mask template for x direction, G y The template is masked for the y-direction,k is a neighborhood marking matrix. According to the template, calculating the amplitude and the direction of the image gradient as follows:
the edges of the image in four directions of horizontal, vertical, diagonal and the like can be detected by the above formula. The direction of its maximum and gradient is identified at each pixel so that we generate a gradient map for each pixel from the original image.
43 Searching for a local maximum of pixels: to determine if a point is an edge point, a local optimum placement may be used, with this method most non-edge points can be removed.
As shown in fig. 2, a schematic diagram of determining a local maximum value of a pixel provided in the embodiment of the present application is shown, and 8 gray values in the neighborhood direction around a point c are found with the point c as the center, and then whether the point c is the largest point is determined by comparison. a. The b direction is the gradient direction of point c and is also where its local maximum is located. And comparing the gray value of the point c with the gray value of the intersection points a and b, and judging that the point c is the point with the maximum local gray value if the gray value of the point c is the maximum.
44 Dual thresholding to solve for image edges: setting a high threshold a and a low threshold b, and if the amplitude of a pixel is larger than a, determining the pixel as an edge pixel, and reserving the edge pixel; if the amplitude of a certain pixel is smaller than b, the pixel is not an edge pixel, and the pixel is removed; if a pixel has an amplitude between a and b, it can be considered an edge pixel only if the pixel is interconnected with a pixel higher than a, leaving it.
After the four steps, each edge contour in the image of the target tire can be extracted, wherein the edge contour of the target tire is included, and finally, a shape characteristic diagram is obtained.
It should be noted that each extracted feature map is renormalized to [0, 255] to ensure fair competition of each feature. Correspondingly, after the four feature maps are extracted, a color feature saliency map, a texture feature saliency map, a brightness feature saliency map and a shape feature saliency map are correspondingly generated by processing each feature map.
According to the tire ground contact area measurement method based on image feature extraction, according to each feature saliency map and the corresponding normalized optimal weight value, the obtaining of the saliency map of the image may include:
obtaining a saliency map of the image using i=g (C) c+g (T) t+g (B) b+g (S) S;
wherein I is a saliency map of an image, C is a color feature saliency map, g (C) is an optimal weight value after normalization processing corresponding to the color feature saliency map, T is a texture feature saliency map, g (T) is an optimal weight value after normalization processing corresponding to the texture feature saliency map, B is a luminance feature saliency map, g (B) is an optimal weight value after normalization processing corresponding to the luminance feature saliency map, S is a shape feature saliency map, and g (S) is an optimal weight value after normalization processing corresponding to the shape feature saliency map.
In the present application, i=g (C) c+g (T) t+g (B) b+g (S) S may be specifically used to obtain a saliency map I of an image of the target tire, where C is a color feature saliency map, g (C) is an optimal weight value after normalization processing corresponding to the color feature saliency map,c. t, b and s are respectively the optimal weight values corresponding to the color feature map, the texture feature map, the brightness feature map and the shape feature map, T is the texture feature saliency map, g (T) is the normalized optimal weight value corresponding to the texture feature saliency map >B is a brightness characteristic saliency map, g (B) is an optimal weight value after normalization processing corresponding to the brightness characteristic saliency map, and +.>S is a shape feature saliency map, g (S) is an optimal weight value after normalization processing corresponding to the shape feature saliency map, and +.>
Through the process, the salient image of the target tire can be obtained, so that the accuracy of the measurement of the ground contact area of the tire is improved.
The method for measuring the tire ground contact area based on image feature extraction provided by the embodiment of the application, according to the target tire in the image, calculates the optimal weight value of each feature saliency map, may include:
and calculating the optimal weight value of each characteristic saliency map by using a genetic algorithm according to the target tire in the image.
In the application, the optimal weight value of each feature saliency map can be calculated by utilizing a genetic algorithm according to the target tire in the image of the target tire, so that the most accurate region of interest can be obtained according to the optimal weight value of each feature saliency map.
The method for measuring the tire ground contact area based on image feature extraction provided by the embodiment of the application, according to the target tire in the image, calculates the optimal weight value of each feature saliency map by using a genetic algorithm, may include:
Randomly selecting a first preset number of feature weights as a first generation of individuals for each feature saliency map;
calculating a feature ROI image under each individual according to each individual and the feature saliency map, and comparing the feature ROI image under each individual with the similarity of the region of interest of the image; the similarity of the region of interest is obtained by multiplying the point-to-point line similarity and the position similarity;
selecting two feature weights corresponding to the feature ROI images with highest image similarity as target feature weights, and judging whether the two target feature weights meet a set condition or not;
if not, randomly selecting a second preset number of feature weights between the two target feature weights, taking the two target feature weights and the second preset number of feature weights as next generation individuals, and executing the step of calculating a feature ROI image under each individual according to each individual and the feature saliency map; the first preset number is different from the second preset number by two;
if yes, calculating the average value of the two target feature weights, and taking the average value of the two target feature weights as the optimal weight value of the feature saliency map.
In the application, after the four features are the four features and the four feature saliency maps are correspondingly obtained after being processed, for the four feature saliency maps, namely the color feature saliency map, the texture feature saliency map, the brightness feature saliency map and the shape feature saliency map, the color feature saliency map and the texture feature saliency map are more important to the tire region image extraction through relevant experiments, so that the optimal weight values of the color feature saliency map and the texture feature saliency map can be obtained by utilizing a genetic algorithm first, and then the optimal weight values of the brightness feature saliency map and the shape feature saliency map can be obtained.
In addition, for the color feature saliency map, the process of calculating the corresponding optimal weight value by using a genetic algorithm specifically comprises the following steps:
(1) Randomly selecting a first preset number of feature weights as first-generation individuals, and multiplying each selected feature weight by the color feature saliency map according to each individual to extract a feature ROI (region of interest) image under each individual (namely, a feature ROI image under each feature weight);
(2) The feature ROI image under each individual is compared with the original image (i.e., the image of the target tire) for region of interest similarity. The similarity of the region of interest is obtained by multiplying the point-to-point similarity and the position similarity, and the region of interest is a region corresponding to the target tire;
(3) Screening out two feature weights (namely individuals) corresponding to the feature ROI images with highest image similarity, and taking the two selected feature weights as target feature weights;
(4) Judging whether the two target feature weights meet the set conditions or not; the setting condition of the method specifically refers to whether the target feature weight reaches the two digits after the decimal point;
(5) If the two target feature weights do not meet the set condition, a second preset number (the second preset number is added with two being equal to the first preset number) of feature weights and the two feature weights are taken as a first preset number of individuals of the next generation between the two target feature weights, then, the step of multiplying each selected feature weight by a color feature saliency map according to each individual to extract a feature ROI image under each individual is performed, namely, the step of multiplying each selected feature weight by the color feature saliency map according to each individual in the step of returning to the step of executing (1) to extract a feature ROI (region of interest) image under each individual is performed;
(6) If the two selected target feature weights meet the set conditions, calculating the average value of the two selected target feature weights, and taking the average value of the two target feature weights as the optimal weight value of the color feature saliency map.
It should be noted that the first preset number mentioned above is greater than 2, and may be specifically 5, or of course, may be other numbers, and in addition, the solving process of the optimal weight values of the three other feature saliency maps is similar to that of the color feature saliency maps, which is not described herein again.
The embodiment of the application also provides a tire ground contact area measuring device based on image feature extraction, referring to fig. 3, which shows a schematic structural diagram of the tire ground contact area measuring device based on image feature extraction, which may include:
an acquiring module 31, configured to acquire an image of a target tire, extract a plurality of feature maps from the image, and process each feature map to generate a corresponding feature saliency map;
a first calculation module 32, configured to calculate an optimal weight value of each feature saliency map according to a target tire in the image, and perform normalization processing on each optimal weight value;
The tire area image obtaining module 33 is configured to obtain a significant image of an image according to each feature significant image and the corresponding normalized optimal weight value, and perform threshold segmentation on the significant image of the image to obtain a tire area image;
the second calculation module 34 is configured to obtain a width of the target tire and a contact length of the target tire with the ground according to the tire area image, and calculate a ground contact area of the target tire according to the width and the contact length.
The tire ground contact area measuring device based on image feature extraction provided in the embodiment of the present application, the second calculating module 34 may include:
the first acquisition unit is used for identifying the specification mark of the target tire, the number of pixels corresponding to the rim diameter and the number of grounding pixels of the target tire from the tire area image, and acquiring the rim diameter from the specification mark;
the first calculating unit is used for calculating the scale factor of the image by utilizing the rim diameter and the number of pixels corresponding to the rim diameter;
and the second calculating unit is used for calculating the contact length of the target tire and the ground by using the number of the ground contact pixels of the target tire and the scaling factor of the image.
The tire ground contact area measuring device based on image feature extraction provided in the embodiment of the present application, the obtaining module 31 may include:
And the processing unit is used for processing each characteristic map through the Gaussian pyramid and the central peripheral difference operator and generating a characteristic saliency map corresponding to each characteristic map.
The tire ground contact area measuring device based on image feature extraction provided in the embodiment of the present application, the obtaining module 31 may include:
and the extraction unit is used for extracting the color feature map, the texture feature map, the brightness feature map and the shape feature map from the image.
The tire ground contact area measuring device based on image feature extraction provided in the embodiment of the present application, the tire area image obtaining module 33 may include:
a saliency map unit for obtaining a tire area image using i=g (C) c+g (T) t+g (B) b+g (S) S;
wherein I is a saliency map of an image, C is a color feature saliency map, g (C) is an optimal weight value after normalization processing corresponding to the color feature saliency map, T is a texture feature saliency map, g (T) is an optimal weight value after normalization processing corresponding to the texture feature saliency map, B is a luminance feature saliency map, g (B) is an optimal weight value after normalization processing corresponding to the luminance feature saliency map, S is a shape feature saliency map, and g (S) is an optimal weight value after normalization processing corresponding to the shape feature saliency map.
The embodiment of the present application provides a tire ground contact area measurement apparatus based on image feature extraction, and the first calculation module 32 may include:
and a third calculation unit for calculating the optimal weight value of each characteristic saliency map by using a genetic algorithm according to the target tire in the image.
The tire ground contact area measurement apparatus based on image feature extraction provided in the embodiment of the present application, the third calculation unit may include:
the first selecting subunit is used for randomly selecting a first preset number of feature weights as a first generation of individuals for each feature saliency map;
the first calculating subunit is used for calculating a characteristic ROI image under each individual according to each individual and the characteristic saliency map and comparing the characteristic ROI image under each individual with the image in a similarity of a region of interest; the similarity of the region of interest is obtained by multiplying the point-to-point line similarity and the position similarity;
the judging subunit is used for selecting two feature weights corresponding to the feature ROI image with the highest image similarity as target feature weights and judging whether the two target feature weights meet the set conditions or not;
the second selecting subunit is configured to randomly select a second preset number of feature weights between the two target feature weights if the two target feature weights do not meet the set condition, take the two target feature weights and the second preset number of feature weights as next generation individuals, and execute the step of calculating a feature ROI image under each individual according to each individual and the feature saliency map; the first preset number is different from the second preset number by two;
And the second calculating subunit is used for calculating the average value of the two target feature weights if the two target feature weights meet the set condition, and taking the average value of the two target feature weights as the optimal weight value of the feature saliency map.
The embodiment of the application also provides a tire ground contact area measurement device based on image feature extraction, referring to fig. 4, which shows a schematic structural diagram of the tire ground contact area measurement device based on image feature extraction provided in the embodiment of the application, may include:
a memory 41 for storing a computer program;
the processor 42, when executing the computer program stored in the memory 41, may implement the following steps:
acquiring an image of a target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map; calculating the optimal weight value of each characteristic saliency map according to the target tire in the image, and carrying out normalization processing on each optimal weight value; obtaining a salient image of the image according to each characteristic salient image and the corresponding normalized optimal weight value, and carrying out threshold segmentation on the salient image of the image to obtain a tire area image; and acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length.
The embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps can be realized:
acquiring an image of a target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map; calculating the optimal weight value of each characteristic saliency map according to the target tire in the image, and carrying out normalization processing on each optimal weight value; obtaining a salient image of the image according to each characteristic salient image and the corresponding normalized optimal weight value, and carrying out threshold segmentation on the salient image of the image to obtain a tire area image; and acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length.
The computer readable storage medium may include: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The description of the relevant parts in the tire grounding area measuring device, the device and the computer readable storage medium based on the image feature extraction provided in the embodiments of the present application may refer to the detailed description of the corresponding parts in the tire grounding area measuring method based on the image feature extraction provided in the embodiments of the present application, which is not repeated herein.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements is inherent to. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element. In addition, the parts of the above technical solutions provided in the embodiments of the present application, which are consistent with the implementation principles of the corresponding technical solutions in the prior art, are not described in detail, so that redundant descriptions are avoided.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (5)

1. The tire ground contact area measuring method based on image feature extraction is characterized by comprising the following steps of:
acquiring an image of a target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map;
calculating the optimal weight value of each characteristic saliency map according to the target tire in the image, and carrying out normalization processing on each optimal weight value;
obtaining a salient map of the image according to each characteristic salient map and the corresponding normalized optimal weight value, and performing threshold segmentation on the salient map of the image to obtain a tire area image;
acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length;
processing each feature map to generate a corresponding feature saliency map, including:
processing each feature map through a Gaussian pyramid and a central peripheral difference operator to generate a feature saliency map corresponding to each feature map;
extracting a plurality of feature maps from the image, including:
Extracting a color feature map, a texture feature map, a brightness feature map and a shape feature map from the image;
obtaining a saliency map of the image according to each characteristic saliency map and the corresponding normalized optimal weight value, wherein the saliency map comprises:
obtaining a saliency map of the image using i=g (C) c+g (T) t+g (B) b+g (S) S; wherein I is a saliency map of the image, C is a color feature saliency map, g (C) is an optimal weight value after normalization processing corresponding to the color feature saliency map, T is a texture feature saliency map, g (T) is an optimal weight value after normalization processing corresponding to the texture feature saliency map, B is a brightness feature saliency map, g (B) is an optimal weight value after normalization processing corresponding to the brightness feature saliency map, S is a shape feature saliency map, and g (S) is an optimal weight value after normalization processing corresponding to the shape feature saliency map;
calculating an optimal weight value of each characteristic saliency map according to a target tire in the image, wherein the optimal weight value comprises the following steps:
randomly selecting a first preset number of feature weights as a first generation of individuals for each feature saliency map; calculating a characteristic ROI image under each individual according to each individual and the characteristic saliency map, and comparing the characteristic ROI image under each individual with the image in a similarity mode; the similarity of the region of interest is obtained by multiplying the point-to-point line similarity and the position similarity; selecting two feature weights corresponding to the feature ROI images with the highest image similarity as target feature weights, and judging whether the two target feature weights meet a set condition or not; if not, randomly selecting a second preset number of feature weights between the two target feature weights, taking the two target feature weights and the second preset number of feature weights as next generation individuals, and executing the step of calculating a feature ROI image under each individual according to each individual and the feature saliency map; wherein the first preset number differs from the second preset number by two; if yes, calculating the average value of the two target feature weights, and taking the average value of the two target feature weights as the optimal weight value of the feature saliency map.
2. The method for measuring the tire contact area based on the image feature extraction according to claim 1, wherein acquiring the contact length of the target tire with the ground from the tire area image comprises:
identifying a specification mark of the target tire, the number of pixels corresponding to the rim diameter and the number of grounding pixels of the target tire from the tire area image, and acquiring the rim diameter from the specification mark;
calculating a scale factor of the image by using the rim diameter and the number of pixels corresponding to the rim diameter;
and calculating the contact length of the target tire and the ground by using the number of the grounding pixels of the target tire and the scale factor of the image.
3. A tire ground contact area measuring device based on image feature extraction, characterized by comprising:
the acquisition module is used for acquiring an image of the target tire, extracting a plurality of feature images from the image, and processing each feature image to generate a corresponding feature saliency map;
the first calculation module is used for calculating the optimal weight value of each characteristic significant figure according to the target tire in the image and carrying out normalization processing on each optimal weight value;
The tire region image obtaining module is used for obtaining a significant image of the image according to each characteristic significant image and the corresponding normalized optimal weight value, and carrying out threshold segmentation on the significant image of the image to obtain a tire region image;
the second calculation module is used for acquiring the width of the target tire and the contact length of the target tire with the ground according to the tire area image, and calculating the grounding area of the target tire according to the width and the contact length;
wherein, acquire the module, include:
the processing unit is used for processing each characteristic map through a Gaussian pyramid and a central peripheral difference operator and generating a characteristic significant map corresponding to each characteristic map;
an acquisition module comprising:
an extracting unit for extracting a color feature map, a texture feature map, a brightness feature map and a shape feature map from the image;
obtaining a tire area image module comprising:
a saliency map obtaining unit for obtaining a saliency map of the image using i=g (C) c+g (T) t+g (B) b+g (S) S; wherein I is a saliency map of the image, C is a color feature saliency map, g (C) is an optimal weight value after normalization processing corresponding to the color feature saliency map, T is a texture feature saliency map, g (T) is an optimal weight value after normalization processing corresponding to the texture feature saliency map, B is a brightness feature saliency map, g (B) is an optimal weight value after normalization processing corresponding to the brightness feature saliency map, S is a shape feature saliency map, and g (S) is an optimal weight value after normalization processing corresponding to the shape feature saliency map;
A first computing module comprising:
the first selecting subunit is used for randomly selecting a first preset number of feature weights as a first generation of individuals for each feature saliency map;
a first calculating subunit, configured to calculate a feature ROI image under each individual according to each individual and the feature saliency map, and compare a region of interest similarity between the feature ROI image under each individual and the image; the similarity of the region of interest is obtained by multiplying the point-to-point line similarity and the position similarity;
the judging subunit is used for selecting two feature weights corresponding to the feature ROI image with the highest image similarity as target feature weights and judging whether the two target feature weights meet a set condition or not;
the second selecting subunit is configured to randomly select a second preset number of feature weights between the two target feature weights if not, take the two target feature weights and the second preset number of feature weights as next generation individuals, and execute the step of calculating a feature ROI image under each individual according to each individual and the feature saliency map; wherein the first preset number differs from the second preset number by two;
And the second calculating subunit is used for calculating the average value of the two target feature weights if yes, and taking the average value of the two target feature weights as the optimal weight value of the feature saliency map.
4. A tire ground contact area measurement apparatus based on image feature extraction, comprising:
a memory for storing a computer program;
a processor for implementing the steps of the tire footprint measurement method based on image feature extraction as claimed in claim 1 or 2 when executing said computer program.
5. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, implements the steps of the tyre ground contact area measurement method based on image feature extraction as claimed in claim 1 or 2.
CN202110003922.8A 2021-01-04 2021-01-04 Tire ground contact area measuring method and device based on image feature extraction Active CN112767471B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110003922.8A CN112767471B (en) 2021-01-04 2021-01-04 Tire ground contact area measuring method and device based on image feature extraction

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110003922.8A CN112767471B (en) 2021-01-04 2021-01-04 Tire ground contact area measuring method and device based on image feature extraction

Publications (2)

Publication Number Publication Date
CN112767471A CN112767471A (en) 2021-05-07
CN112767471B true CN112767471B (en) 2024-03-29

Family

ID=75699637

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110003922.8A Active CN112767471B (en) 2021-01-04 2021-01-04 Tire ground contact area measuring method and device based on image feature extraction

Country Status (1)

Country Link
CN (1) CN112767471B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009122939A (en) * 2007-11-14 2009-06-04 Bridgestone Corp Feature extraction program creation device for tire inspection and tire inspection device
CN103279957A (en) * 2013-05-31 2013-09-04 北京师范大学 Method for extracting remote sensing image interesting area based on multi-scale feature fusion
JP2017027600A (en) * 2015-07-23 2017-02-02 株式会社リコー Gesture identifying method and device
CN111735524A (en) * 2020-08-27 2020-10-02 湖南大学 Tire load obtaining method based on image recognition, vehicle weighing method and system
KR20200124827A (en) * 2019-04-25 2020-11-04 인제대학교 산학협력단 Alzheimer's disease classification based on multi-feature fusion

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009122939A (en) * 2007-11-14 2009-06-04 Bridgestone Corp Feature extraction program creation device for tire inspection and tire inspection device
CN103279957A (en) * 2013-05-31 2013-09-04 北京师范大学 Method for extracting remote sensing image interesting area based on multi-scale feature fusion
JP2017027600A (en) * 2015-07-23 2017-02-02 株式会社リコー Gesture identifying method and device
KR20200124827A (en) * 2019-04-25 2020-11-04 인제대학교 산학협력단 Alzheimer's disease classification based on multi-feature fusion
CN111735524A (en) * 2020-08-27 2020-10-02 湖南大学 Tire load obtaining method based on image recognition, vehicle weighing method and system

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于Faster R-CNN与形态法的路面病害识别;晏班夫;徐观亚;栾健;林杜;邓露;中国公路学报;20210930;第34卷(第9期);第【0181】-【0193】页 *

Also Published As

Publication number Publication date
CN112767471A (en) 2021-05-07

Similar Documents

Publication Publication Date Title
US9286533B2 (en) Method for image recombination of a plurality of images and image identification and system for image acquiring and identification
US6674915B1 (en) Descriptors adjustment when using steerable pyramid to extract features for content based search
CN104778721A (en) Distance measuring method of significant target in binocular image
CN104318225A (en) License plate detection method and device
CN111382658B (en) Road traffic sign detection method in natural environment based on image gray gradient consistency
CN116542982B (en) Departure judgment device defect detection method and device based on machine vision
CN107767387A (en) Profile testing method based on the global modulation of changeable reception field yardstick
CN111931709A (en) Water body extraction method and device for remote sensing image, electronic equipment and storage medium
CN105139011A (en) Method and apparatus for identifying vehicle based on identification marker image
CN113781413B (en) Electrolytic capacitor positioning method based on Hough gradient method
CN112767471B (en) Tire ground contact area measuring method and device based on image feature extraction
CN108520252B (en) Road sign identification method based on generalized Hough transform and wavelet transform
CN113252103A (en) Method for calculating volume and mass of material pile based on MATLAB image recognition technology
CN112633070A (en) High-resolution remote sensing image building extraction method and system
CN110008955B (en) Method for testing character imprinting quality of surface of automobile brake pad
CN115100510B (en) Tire wear degree identification method
CN111144413A (en) Iris positioning method and computer readable storage medium
CN111476233A (en) License plate number positioning method and device
CN113420633B (en) Traffic sign identification method based on UM enhancement and SIFT feature extraction
JP2611723B2 (en) Image sharpness measuring device
Joshi et al. A simple scheme for contour detection.
CN113128500A (en) Mask-RCNN-based non-motor vehicle license plate recognition method and system
CN112329556A (en) Traffic sign detection method and device, electronic equipment and storage medium
CN112785550A (en) Image quality value determination method, image quality value determination device, storage medium, and electronic device
JP3358997B2 (en) Engraved mark identification device

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