Footprint Pressure Feature Analysis System Based on Segmentation Region Method

Abstract

The present invention discloses a footprint pressure feature analysis system based on a segmentation region method, which relates to the technical field of footprint analysis and solves the technical problem in the traditional solution that during the extraction of footprint pressure features, instability is brought about in the traditional manual analysis process, resulting in low efficiency and accuracy in the extraction of footprint pressure information. The present invention performs pressure analysis on the collected footprint images, and quantitatively analyzes the pressure information hidden in the footprint images through a small amount of calculations. It can not only calculate the heavy pressure distribution of the footprint through the distribution of image gray values, but also obtain the pressure surface information and footprint pressure lines through the footprint images. Compared with the traditional method of manual analysis, it greatly speeds up the analysis speed, simplifies the analysis process, improves the analysis accuracy, has high application prospects in the fields of public safety, etc., and adapts to the general trend of intelligent development.

Description

Technical Field

[0001] The present invention belongs to the field of footprint analysis, relates to footprint pressure feature analysis technology, and specifically is a footprint pressure feature analysis system based on a segmentation region method. Background Art

[0002] With the rapid development of biometric recognition technology, footprint pressure information, as one of the important biometric features, is being further developed and studied. Footprint pressure features are widely used in many occasions, but there are still many problems in the extraction of footprint pressure features.

[0003] At the scene, there are often footprint information left by criminal suspects. These footprint information often appear in gray form, and it is impossible to clearly obtain the pressure information therein; during the comparison process of on-site footprints, there are often situations that cannot be accurately distinguished by the naked eye, which greatly reduces the efficiency and accuracy of handling cases; therefore, there is an urgent need for a feature analysis system that can accurately and efficiently extract footprint pressure information. Summary of the Invention

[0004] The present invention provides a footprint pressure feature analysis system based on a segmentation region method, which is used to solve the technical problems that in the traditional solution, when extracting footprint pressure features, instability is brought in the traditional manual analysis process, resulting in low extraction efficiency and extraction accuracy of footprint pressure information; the present invention provides an algorithm program for automatically analyzing footprint pressure information, footprint pressure distribution and extracting footprint pressure lines, and integrates the three algorithms. The operator only needs a few simple steps to obtain the pressure information hidden in the footprint image; it provides convenience for experts to analyze footprint pressure, can greatly improve the speed and accuracy of footprint pressure information analysis at the scene, and also provides convenience for subsequent footprint information comparison work.

[0005] The object of the present invention can be achieved through the following technical solutions: A footprint pressure feature analysis system based on a segmentation region method, including a data acquisition module, a pressure surface drawing module, a sole region segmentation module, and a footprint pressure line drawing module;

[0006] The data acquisition module acquires a footprint image and sends the footprint image to the pressure surface drawing module;

[0007] The pressure surface drawing module is used to perform conversion processing on the footprint image to obtain a footprint pressure image and footprint biological parameters, and send the footprint pressure image and footprint biological parameters to the sole region segmentation module; wherein, the conversion processing includes footprint pressure information conversion and image denoising;

[0008] The sole area segmentation module segments the footprint to obtain the footprint area and determines the footprint center line; wherein, the footprint area includes a toe area, a sole area, and a heel area, and the footprint biological parameters include the upper vertex of the footprint, the lower vertex of the footprint, the center point of the sole, the center point of the heel, and the corresponding coordinates.

[0009] The footprint pressure line drawing module obtains the footprint pressure line and the footprint pressure function according to the footprint pressure image and the footprint area.

[0010] Preferably, the footprint image is obtained by a footprint scanner.

[0011] Preferably, the footprint pressure information conversion includes background normalization and image filtering, and the image filtering includes median filtering; wherein, the background normalization specifically includes:

[0012] Calculate the gray average value of the pixel points of the footprint image, set the background of the footprint image according to the gray average value, and obtain the gray histogram of the footprint image.

[0013] Preferably, an initial threshold is obtained according to the gray histogram of the footprint image, and gray segmentation is performed on the footprint image histogram through the initial threshold to obtain color levels, and different color levels are colored.

[0014] Preferably, the image denoising includes manual denoising and automatic denoising; wherein, the automatic denoising selects the main area by means of frame selection and denoises the part outside the main area; the main area is the footprint area in the footprint pressure image.

[0015] Preferably, the footprint length in the footprint pressure image is obtained according to the footprint biological parameters, and the footprint length is combined with the overweight ratio and the constraint condition to segment the footprint pressure image to obtain the footprint area and the corresponding length; wherein, the overweight ratio is the ratio of the toes, sole, arch, and heel of a normal person to the entire sole.

[0016] Preferably, the constraint condition is specifically that the sum of the ratios of the toes, sole, arch, and heel to the entire sole is 1.

[0017] Preferably, the footprint center line is generated by fitting the coordinates of the center point of the sole and the center point of the heel.

[0018] Preferably, after obtaining the footprint center, it is also necessary to calculate the overweight ratio, including:

[0019] Obtain the number of pressure points on both sides of the footprint center line in combination with the footprint pressure image; calculating the heavy pressure ratio is part of the region segmentation module, and it is calculated based on the result of region segmentation and the number of pressure points on both sides of the center line. Among them, the calculation method of the number of pressure points is to count the number of effective pixel points on both sides of the center line of the image after pressure conversion (in the RGB three channels: pixel points with an R channel of 200 or more and G and B channels of 20 or less), and one pixel point is one heavy pressure point.

[0020] Preferably, after obtaining the footprint area, it is also necessary to determine the upper and lower boundaries corresponding to the footprint area, including:

[0021] Take the ordinate of the upper vertex of the footprint as the upper boundary of the toe area, and combine the length of the toe area to obtain the lower boundary of the toe area, and the lower boundary of the toe area is the upper boundary of the sole area;

[0022] The upper boundary of the sole area is combined with the sole length to obtain the lower boundary of the sole area;

[0023] Take the lower vertex of the footprint as the lower boundary of the heel area, and the lower boundary of the heel area is combined with the length of the heel area to obtain the upper boundary of the heel area.

[0024] Preferably, after obtaining the footprint area, it is also necessary to determine the area width corresponding to the footprint area, including:

[0025] Scan each row of pixel points in the footprint area one by one, determine the abscissa values of the two pixel points with the largest difference in abscissa within the corresponding footprint area, and use the abscissa values of these two pixel points as the left and right boundaries of the corresponding footprint area.

[0026] Compared with the prior art, the beneficial effects of the present invention are:

[0027] The present invention quantitatively analyzes the pressure information hidden in the footprint image through a small amount of calculations by performing pressure analysis on the collected footprint image. It can not only calculate the heavy pressure distribution of the footprint through the distribution of image gray values, but also obtain the pressure surface information and the footprint pressure line from the footprint image; compared with the traditional method of manual analysis, it greatly speeds up the analysis speed, simplifies the analysis process, improves the analysis accuracy, has a high application prospect in the field of public safety, etc., and adapts to the general trend of intelligent development. Description of the Drawings

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0029] Figure 1 It is a schematic diagram of the principle flow of the present invention;

[0030] Figure 2 It is the footprint image collected in the embodiment of the present invention;

[0031] Figure 3 It is a schematic diagram of the pressure surface drawing process in the embodiment of the present invention;

[0032] Figure 4 It is a schematic diagram of the comparison of different thresholds of the footprint pressure image in the embodiment of the present invention;

[0033] Figure 5 It is a schematic diagram of the region segmentation process in the embodiment of the present invention;

[0034] Figure 6 It is a schematic diagram of the footprint pressure line drawing process in the embodiment of the present invention. Detailed implementation manners

[0035] The following will clearly and completely describe the technical solutions of the present invention in combination with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0036] The terms used here are for describing the embodiments and do not intend to limit and / or restrict the present disclosure; it should be noted that unless the context clearly indicates otherwise, the singular forms of "a", "an" and "the" also include the plural forms; moreover, although terms such as "first", "second", etc. can be used herein to describe various elements, the elements are not limited by these terms, and these terms are only used to distinguish one element from another.

[0037] Please refer to Figure 1 - Figure 2 , the present invention provides a footprint pressure feature analysis system based on the region segmentation method, including a data acquisition module, a pressure surface drawing module, a plantar region segmentation module, and a footprint pressure line drawing module;

[0038] The footprint image is collected by the data acquisition module and sent to the pressure surface drawing module;

[0039] The pressure surface drawing module is used to perform conversion processing on the footprint image to obtain the footprint pressure image and footprint biological parameters, and send the footprint pressure image and footprint biological parameters to the plantar area segmentation module; among them, the conversion processing includes footprint pressure information conversion and image denoising;

[0040] The plantar area segmentation module segments the footprint to obtain the footprint area and determines the footprint center line; among them, the footprint area includes the toe area, the sole area and the heel area, and the footprint biological parameters include the upper vertex of the footprint, the lower vertex of the footprint, the center point of the sole, the center point of the heel and the corresponding coordinates;

[0041] The footprint pressure line drawing module obtains the footprint pressure line and the footprint pressure function according to the footprint pressure image and the footprint area.

[0042] The present invention mainly processes the on-site footprint; first, for the footprint existing on the site, the staff uses a professional footprint scanner to scan and extract to obtain the footprint image, and then further processes the footprint image through the present invention.

[0043] Please refer to Figure 3 - Figure 4 , in the present invention, the pressure surface drawing module is used to convert the footprint image into a footprint pressure image, and the conversion process includes footprint pressure information conversion and image denoising.

[0044] The footprint pressure information conversion consists of two parts: background normalization and image filtering; the image filtering method in the present invention adopts median filtering.

[0045] Background normalization mainly considers that for different acquisition devices or acquisition sites, the background of the footprint image may be inconsistent; this process determines the background color of the image by calculating the average value of the pixel points of the whole image; generally speaking, the average value of the image pixels will tend to the color of the image background, so for the convenience of subsequent operations, the present invention sets the background of all input footprint images to black and the footprint itself to white.

[0046] In the present invention, the background normalization specifically includes:

[0047] Calculate the gray average value of the pixel points of the footprint image, set the background of the footprint image according to the gray average value, and obtain the gray histogram of the footprint image.

[0048] In the footprint pressure feature analysis system based on the segmentation region method provided by the present invention, an initial threshold is obtained through the gray histogram of the footprint image, and gray segmentation is performed on the footprint image histogram through the initial threshold to obtain color levels, and different color levels are colored.

[0049] By traversing, calculate the number of occurrences of each gray value in the gray histogram of the footprint image, and then adaptively set an initial threshold according to the distribution of the gray histogram. The purpose of this initial value is to segment the color gradient. The present invention combines the gray histogram of the footprint image and formula (1) to calculate the adaptive initial threshold;

[0050]

[0051] Among them, K th is the initial threshold, X i is the number of occurrences of the corresponding gray value i, and M represents the gray value corresponding to this number. The function of the initial threshold is to divide the gray levels of the gray histogram of the footprint image into 5 parts, and the number of pixel points in each gray level is kept consistent.

[0052] Coloring is for different pressure colors on different color scales, referring to formula (2), formula (3) and formula (4).

[0053]

[0054]

[0055]

[0056] Among them, img B (i, j), img G (i, j), img R (i, j) are the pixel values of each channel of the BGR of the obtained footprint pressure image, and G(i, j) is the pixel value of the gray histogram of the footprint image; x is the gray color difference obtained through the initial threshold, and the purpose is to evenly distribute different pressure colors to areas with different gray values, and is obtained through formula 5; v is the value of the "color enhancement" slider.

[0057]

[0058] For example: when the color scale of a certain point in the gray histogram of the footprint image is located in (0, K th ), it is considered that the pressure value of this point is very small, and the color with BGR value (0, 0, 0) (black) will be assigned to this point; when the color scale of a certain point in the gray histogram of the footprint image is located in (K th + 3x, K th + 4x), it is considered that the pressure value of this point is very large, and the color with BGR value (0, 255 - v, 255) (red) will be assigned to this point.

[0059] Please refer to Figure 4“Color enhancement” controls the pressure color presented in the image by adjusting the value of the slider. When the slider value is lower, the color distribution of the pressure image is wider, as shown in Figure 4 (a); if the slider value is high, the color distribution of the pressure image will be narrower, as shown in Figure 4 (b); “Background segmentation” controls the number of coloring points to achieve background segmentation, that is, only the points in the image that exceed the threshold will participate in the coloring algorithm, and those below the threshold will be directly determined as black, as shown in Figure 4 (c).

[0060] The image denoising in the present invention includes two methods: manual denoising and automatic denoising, which are used to remove some difficult-to-remove noises in the footprint pressure image; manual denoising is to manually erase the existing noises in the image by using an eraser; automatic denoising is to select the main area of the footprint pressure image by means of frame selection, and then perform regional denoising on the part outside the main area. The stubborn noise information existing in the image can be removed by combining the two denoising methods; finally, click the “Save” button to save the footprint pressure image in the current state.

[0061] It is worth mentioning that algorithms for footprint correction and footprint key point extraction are also embedded in the save button. Through these algorithms, a corrected footprint pressure image can be obtained, as well as the coordinates of the four points: the upper vertex, the lower vertex, the center point of the sole, and the center point of the heel of the footprint; the algorithms for footprint correction and footprint key point extraction will not be introduced in detail here.

[0062] Please refer to Figure 5 In a footprint pressure feature analysis system based on the segmentation region method provided by the present invention, the sole region segmentation module segments the footprint to obtain the footprint region and determines the footprint center line.

[0063] The sole region segmentation module calculates the weight ratio of the footprint pressure through the footprint pressure image and automatically divides the footprint into three regions: toes, sole, and heel. The present invention adds the width information of the footprint region compared with the traditional method.

[0064] Before the sole region segmentation, it is necessary to extract the heavy pressure region, that is, all low gray value pixel points are set to (0, 0, 0), and only high gray value pixel points are retained, so that effective foreground segmentation can be carried out; a part of the region with a larger pressure value is retained, and the region with a smaller pressure value is removed.

[0065] According to the coordinates of the two points of the upper vertex (x_top, y_top) and the lower vertex (x_bottom, y_bottom) of the footprint, and formula (6), the footprint length is obtained;

[0066] L = y a-y b #(6)

[0067] Where: L is the footprint length; y a is y_bottom, y b is y_top.

[0068] After calculating the footprint length, further calculations are needed to determine the length information of the toe, sole, and heel regions; it can be calculated (refer to the book "Footprint Science") that the proportions of the toe, sole, arch, and heel regions of a normal person in the entire sole are: λ1, λ2, λ3, and λ4 respectively, and satisfy the constraint conditions of formula (7); according to the above proportional relationship and constraint relationship, calculate the specific length information of each region, as shown in formula (8):

[0069]

[0070]

[0071] Where: L t is the length of the toe region, L s is the length of the sole region, L h is the length of the heel region; λ1, λ2, λ3, and λ4 take the values of 0.1589, 0.2322, 0.3389, and 0.2700 respectively.

[0072] Taking y_top as the ordinate of the upper boundary of the toe region, the ordinate of the lower boundary of the toe region is determined as y_top + L t ; then taking the ordinate of the lower boundary of the toe region as the ordinate of the upper boundary of the sole region, the ordinate of the lower boundary of the sole region is further determined as y_top + L t +L s ; finally, since the sole region and the heel region of a person are not directly connected under normal circumstances, the present invention takes y_bottom as the ordinate of the lower boundary of the heel region, and the ordinate of the upper boundary of the heel region is determined as: y_bottom - L h ; thus, the 6 boundary ordinates of the three regions to be sought are determined.

[0073] Finally, for the calculation of the region width, the footprint pressure image is divided into three regions by the ordinates of the three regions, and the width of each region is determined respectively in each region; the present invention determines the abscissa values of the two points with the largest difference in abscissa in the region by scanning each pixel point in each row in the region one by one, and takes the abscissa values of these two points as the left and right boundaries of this region; there are three regions in total, and 6 boundary abscissas are determined in total.

[0074] Through the above calculations, the present invention determines the specific coordinates of 12 points in three regions; it should be noted that the lower boundary of the toe region coincides with the upper boundary of the sole region, which also conforms to the footprint characteristics of normal people.

[0075] In addition, the present invention fits a footprint centerline function through the coordinates of the center point of the front sole (x_cen_top, y_cen_top) and the center point of the heel (x_cen_bottom, y_cen_bottom), and uses this function as the center of the footprint region; next, calculate the number of pressure points on both sides of the centerline in each of the three regions; among them, the calculation method of the number of pressure points is to count the number of effective pixel points on both sides of the image centerline after pressure conversion (in the RGB three channels: pixel points with R channel being 200 or more and G and B channels being 20 or less are effective pixel points), and one effective pixel point is a heavy pressure point; determine the weight bias of the footprint by the number of pressure points on both the left and right sides. In most cases, the more the number of points (the greater the density), the greater the pressure on this part, that is, the footprint pressure is biased towards this step; the heavy pressure ratio is calculated according to formula (9):

[0076]

[0077] Among them, N h represents the number of pressure points in the heavy pressure bias region, N l represents the number of pressure points in the non-pressure bias region, and when N l = 0, the ratio is 1.

[0078] The above process is the main process of the present invention, mainly automatically dividing the three footprint regions of the toe, sole, and heel according to the previously obtained known information, and calculating the footprint pressure bias in the three regions; compared with the traditional footprint region segmentation algorithm, this algorithm more accurately divides the three footprint regions, not only divides the footprint region in height, but also constrains the region of the footprint image in width, so that the finally presented form is a rectangular frame, and the segmentation effect is better; in addition, the determination of the footprint centerline also lays a foundation for calculating the footprint pressure bias.

[0079] Please refer to Figure 6 , in a footprint pressure feature analysis system provided by the present invention based on a segmentation region method, the footprint pressure line drawing module obtains the footprint pressure line and the footprint pressure function according to the footprint pressure image and the footprint region.

[0080] The footprint pressure line drawing module mainly obtains the footprint pressure line based on the footprint pressure image and the result of the footprint segmentation area; as the name implies, the footprint pressure line is the pressure distribution curve of a footprint image. For normal people, the general footprint pressure line is in the shape of a lightning bolt; this curve can clearly represent the distribution of the footprint pressure and provides convenience for footprint comparison.

[0081] For the existing footprint pressure line acquisition algorithms, generally, based on frame-by-frame GIF pictures or footprint grayscale pictures with grayscale information, the footprint pressure line is selected. The color space of this kind of picture is linearly distributed, and the method of finding the point with the largest color pressure value is relatively simple and only requires simple sorting; however, for a color pressure image, the distribution of the color space is non-linear. This algorithm finds a relatively perfect non-linear mapping relationship according to the distribution relationship of the color space of the color pressure image, maps the BGR color space to the grayscale color space in the order of pressure values; then convolution operations are performed to reduce the possible noise in the image, and finally, the operation of connecting lines in different regions is carried out; compared with the existing solutions, the technical solution of the present invention greatly improves the generalization ability of the model and the analysis accuracy.

[0082] The footprint pressure line drawing solution provided by the present invention is as follows:

[0083] Color space compression. This part of the algorithm will score each pixel point in the pressure image according to the color distribution of formulas (2), (3), and (4). The specific scoring rules are shown in formula (10):

[0084]

[0085] Where: S i,j is the pressure score of the current pixel; img B (i, j), img G (i, j), img R (i, j) are the pixel values of the BGR three channels in the pressure image.

[0086] The above scoring method adopts a segmented scoring method, and different scoring rules are used to score the pixel points located in different regions of the color space; the total score is 100 points. The higher the score, the greater the pressure value. It is divided into 5 segments, each segment is 20 points, and addition and subtraction are performed within the score range of this segment according to the specific color difference; such a mapping can ensure that the conversion process from the three-dimensional color space to the one-dimensional space will not be crossed and confused, so that the score after mapping completely conforms to the pressure information of each pixel point in the footprint pressure image.

[0087] This technical solution uses a 3×3 convolution kernel to perform a convolution operation with a step size of 1 and a fill of 1 on the score matrix after color space compression, so that each pixel aggregates the information of other pixels in its 3×3 domain, and then takes the average value. The average value after the aggregation of the pixel is the final score of the pixel. Since there are inevitably some noise points in the footprint image, the advantage of doing this is that it can avoid the possible noise points from affecting the drawing of the footprint pressure line, so that the drawn footprint pressure line has a higher confidence.

[0088] In addition, in order to make the drawn footprint pressure line more stable, combined with the results of regional segmentation, the footprint area is divided into 12 blocks. They are: the toe area is evenly divided into the front toe area and the rear toe area; the sole area is evenly divided into 7 areas because it has more pressure information; the heel area is evenly divided into the heel front edge area, the heel middle area and the heel rear edge area, a total of 12 areas; the present invention will find the area with the highest score in each of these 12 areas as the heavy pressure point of the area. If a certain area has more concentrated high-scoring points, the algorithm will automatically select the center point of the area as the heavy pressure point of the area; finally, a heavy pressure point is selected in each area, a total of 12 heavy pressure points, and they are connected in sequence. The result of the connection is the required footprint pressure line; continue to take out the coordinates of each heavy pressure point, draw it into a rectangular coordinate system with the height of the footprint image as the X-axis and the width as the Y-axis, and then connect them with a broken line to obtain the footprint pressure function.

[0089] A person's footprint pressure line can reflect the person's body characteristics, walking habits and other information. For example, a normal person's footprint pressure line is in the shape of a "lightning bolt". If the person has the habit of walking with one foot, his footprint pressure line will be a straight line. In addition, by drawing the footprint pressure function, quantitative analysis can be performed. By comparing the similarity of two footprint pressure functions, the probability that the footprints are from the same person can be determined.

[0090] The data in the above formula are all calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.

[0091] Working principle of the present invention:

[0092] The footprint image is collected by the data acquisition module and sent to the pressure surface rendering module; the pressure surface rendering module is used to convert the footprint image to obtain the footprint pressure image and footprint biological parameters, and send the footprint pressure image and footprint biological parameters to the plantar area segmentation module.

[0093] The plantar region segmentation module segments the footprint to obtain the footprint region and determines the footprint center line; the footprint pressure line drawing module obtains the footprint pressure line and the footprint pressure function according to the footprint pressure image and the footprint region.

[0094] In the description of this specification, the description referring to terms such as "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0095] The above content is only an example and explanation of the structure of the present invention. Those skilled in the art of this technology can make various modifications or supplements to the specific embodiments described or use similar ways to replace them. As long as they do not deviate from the structure of the application or exceed the scope defined by this claim book, they should all belong to the protection scope of the present invention.

Claims

1. A footprint pressure feature analysis system based on a segmentation region method, characterized in that, It includes a data acquisition module, a pressure surface drawing module, a plantar area segmentation module, and a footprint pressure line drawing module; The data acquisition module acquires a footprint image and sends the footprint image to the pressure surface drawing module; The pressure surface drawing module is used to perform conversion processing on the footprint image to obtain a footprint pressure image and footprint biological parameters, and send the footprint pressure image and footprint biological parameters to the plantar area segmentation module; wherein, the conversion processing includes footprint pressure information conversion and image denoising; The plantar area segmentation module segments the footprint to obtain a footprint area and determines a footprint center line; wherein, the footprint area includes a toe area, a sole area, and a heel area, and the footprint biological parameters include the upper vertex of the footprint, the lower vertex of the footprint, the center point of the sole, the center point of the heel, and the corresponding coordinates; the footprint center line is generated by fitting the coordinates of the center point of the sole and the center point of the heel; The footprint pressure line drawing module obtains a footprint pressure line and a footprint pressure function according to the footprint pressure image and the footprint area; the footprint pressure line drawing scheme is as follows: Color space compression, scoring each pixel point in the pressure image according to the color distribution of formulas (2), (3), and (4): Among them, img B (i, j), img G (i, j), img R (i, j) are the pixel values of each channel of the obtained footprint pressure image in BGR, and G(i, j) are the pixel values of the gray-scale histogram of the footprint image; x is the gray-scale color difference obtained through the initial threshold, and the purpose is to evenly distribute different pressure colors to regions with different gray-scale values; v is the value of the "color enhancement" slider; The specific scoring rule is shown in formula (10): Where: S i,j is the pressure score of the current pixel; img B (i, j), img G (i, j), img R (i, j) are the pixel values of the BGR three channels in the pressure image; A segmented scoring method is adopted, and different scoring rules are used to score pixel points located in different regions of the color space; the total score is 100 points, and the higher the score, the greater the pressure value. It is divided into 5 segments, each segment is 20 points, and addition and subtraction are performed within the score range of the segment according to the difference in specific colors; Use a 3×3 convolution kernel to perform a convolution operation with a stride of 1 and a padding of 1 on the scoring matrix after color space compression, so that each pixel point aggregates the information of other pixel points in its 3×3 neighborhood, and then takes the average value. The average value after aggregation of this pixel point is the final score of this pixel point; The footprint area is divided into 12 blocks; namely: the toe area is evenly divided into a front toe area and a rear toe area; the sole area is evenly divided into 7 areas; the heel area is evenly divided into a front edge area of the heel, a middle area of the heel, and a rear edge area of the heel, totaling 12 areas; for these 12 blocks, find the area with the highest score in each area as the heavy pressure point of this area. When there are many concentrated high-score points in a certain area, the algorithm will automatically select the center point of this area as the heavy pressure point of the area; finally, select one heavy pressure point in each area, totaling 12 heavy pressure points, and connect them in sequence. The connected result is the required footprint pressure line; continue to take out the coordinates of each heavy pressure point, draw them in a rectangular coordinate system with the height of the footprint image as the X-axis and the width as the Y-axis, and then connect them with a broken line to obtain the footprint pressure function.

2. The footprint pressure feature analysis system based on the segmentation region method according to claim 1, characterized in that, The footprint pressure information conversion includes background normalization and image filtering, and the image filtering includes median filtering; wherein, the background normalization specifically includes: Calculate the gray average value of the pixel points of the footprint image, set the background of the footprint image according to the gray average value, and obtain the gray histogram of the footprint image.

3. The footprint pressure feature analysis system based on the segmentation area method according to claim 2, characterized in that, An initial threshold is obtained according to the grayscale histogram of the footprint image. The grayscale segmentation of the footprint image histogram is performed through the initial threshold to obtain color levels, and different segmented color levels are colored.

4. The footprint pressure feature analysis system based on the segmentation region method according to claim 1, wherein The image denoising includes manual denoising and automatic denoising; among them, the automatic denoising selects the main area by means of frame selection, and denoises the part outside the main area; the main area is the footprint area in the footprint pressure image.

5. The footprint pressure characteristic analysis system based on the segmented area method according to claim 1, wherein The footprint length in the footprint pressure image is obtained according to the footprint biological parameters. The footprint length, combined with the overweight ratio and the constraint condition, is used to segment the footprint pressure image to obtain the footprint area and the corresponding length; among them, the overweight ratio is the proportion of the toes, sole, arch, and heel of a normal person in the entire sole.

6. The footprint pressure characteristic analysis system based on the segmentation region method according to claim 5, characterized in that The specific constraint condition is that the sum of the proportions of the toes, sole, arch, and heel in the entire sole is 1.

7. The footprint pressure characteristic analysis system based on the segmentation region method according to claim 1, wherein After obtaining the footprint center line, the overweight ratio needs to be calculated.

8. The footprint pressure characteristic analysis system based on the segmented area method according to claim 1, wherein After obtaining the footprint area, the upper and lower boundaries corresponding to the footprint area need to be determined, including: The ordinate of the upper vertex of the footprint is used as the upper boundary of the toe area, and the lower boundary of the toe area is obtained by combining the length of the toe area, and the lower boundary of the toe area is the upper boundary of the sole area; The upper boundary of the sole area is combined with the sole length to obtain the lower boundary of the sole area; The lower vertex of the footprint is used as the lower boundary of the heel area, and the upper boundary of the heel area is obtained by combining the lower boundary of the heel area with the length of the heel area.

9. The footprint pressure characteristic analysis system based on the segmented area method according to claim 1, characterized in that After obtaining the footprint area, the area width corresponding to the footprint area needs to be determined, including: Each row of pixel points in the footprint area is scanned one by one, and the abscissa values of the two pixel points with the largest difference in abscissa within the corresponding footprint area are determined, and the abscissa values of these two pixel points are used as the left and right boundaries of the corresponding footprint area.

Images