WO2023179465A1

WO2023179465A1 - Image texture extraction method and device, and computer readable storage medium

Info

Publication number: WO2023179465A1
Application number: PCT/CN2023/082070
Authority: WO
Inventors: 张国流
Original assignee: 张国流
Priority date: 2022-03-24
Filing date: 2023-03-17
Publication date: 2023-09-28
Also published as: CN115965671A

Abstract

Disclosed in the present invention are an image texture extraction method and device, and a computer readable storage medium, the method comprising: performing number system conversion on color values of each pixel in a target image to obtain a digital array of each pixel, wherein the base to which the color values of each pixel are converted is less than the base from which the color values of the pixel are converted; according to the number of digits of each digital array, decomposing the target image into a plurality of image layers; performing texture information extraction on the image layers respectively by using a grid mask, so as to obtain a sub-texture map of each image layer; and performing weight superposition on all the sub-texture maps to synthesize an overall texture map of the target image. Compared with the traditional edge detection technique, the present invention can greatly improve the texture information extraction rate and texture information extraction precision for the target image, thus providing more complete texture and edge information for a downstream image information processing procedure.

Description

Image texture extraction method, equipment and computer-readable storage medium

Technical field

The present invention relates to the technical field of image processing, and in particular to an image texture extraction method, equipment and computer-readable storage medium.

Background technique

Image edge and texture extraction play an important role in the fields of human vision and computer vision. Texture extraction (edge detection) refers to the process of using certain technologies and methods to achieve target texture extraction in an image containing a target and background, ignoring the influence of background and noise interference. At present, the traditional texture extraction method uses traditional edge detection operators to detect edge information with significant grayscale changes in the image. Traditional edge detection technology mainly uses grayscale as the detection index, and often needs to convert complex images into grayscale images before performing edge detection processing. Therefore, traditional edge detection technology often outputs monotonous results, serious homogeneity of edge information, and is difficult to Do further subdivision processing; at the same time, because the traditional edge detection algorithm can only respond well to the peak value of grayscale change, it often only has a good detection effect on strong textures, and has a good detection effect on medium or weak textures. Poor, a large amount of weaker but equally important detail information in the image will be lost. All these result in traditional edge detection algorithms performing poorly in terms of completeness and fineness of texture information extraction.

Contents of the invention

In response to the above problems, the purpose of the present invention is to provide an image texture extraction method, device and computer-readable storage medium, which can effectively improve the completeness and precision of image texture extraction. In a first aspect, embodiments of the present invention provide an image texture extraction method, including:

Convert the color value of each pixel of the target image into a base to obtain a digital array for each pixel; where the base of the color value of the pixel after conversion is lower than the base of the color value of the pixel before conversion;

Decompose the target image into several layers according to the digital array;

Use a window frame mask to extract texture information for each layer, and obtain a sub-texture map for each layer;

All the sub-texture maps are superimposed by weights to synthesize a total texture map of the target image.

As an improvement to the above solution, the color value of each pixel of the target image is converted into a binary system to obtain a digital array, including:

Convert the color values of each pixel of the target image into N-ary systems to obtain a digital array for each pixel;

Among them, the color scale values of the R, G, and B color channels of each pixel are expressed in N-base form, and the RGB value of each pixel is expressed as a 3×M-bit digital array; among them, the N-base is smaller than the color value Current forwarding, M represents the number corresponding to the maximum level value of the color channel in the N system.

As an improvement to the above solution, the color values of each pixel of the target image are converted into N numbers to obtain a digital array for each pixel, including:

Perform base conversion on the color values of each pixel of the target image to obtain a 3×M-bit digital array for each pixel;

Then, according to the digital array, the target image is decomposed into several layers, including:

According to the digital array of each pixel, the digits of each color channel in all pixels are clustered and merged into one layer.

As an improvement to the above solution, the window frame mask is used to extract texture information for each layer, and the sub-texture map of each layer is obtained, including:

Each layer is covered with the window frame mask; wherein the window frame mask is composed of several window frames;

For each layer, each window frame uses at least two comparison groups to extract the texture of the layer area where it is located, and obtain the texture information of the layer area where each window frame is located; among them, each comparison group consists of two comparison groups on the window frame. It consists of two detection points on opposite sides of each other;

The texture information of all window frames in the window frame mask is spliced to obtain a sub-texture map of the corresponding layer.

As an improvement to the above solution, a first comparison group and a second comparison group are set on the boundary of each window frame, wherein the line connecting the two detection points of the first comparison group is connected with the two detection points of the second comparison group. The angle between the lines connecting the detection points is 360°/2n, where n represents the number of comparison groups;

Then, each window frame uses at least two pairs of comparison groups to extract the texture of the layer area where it is located, and obtain the texture information of the layer area where each window frame is located, including:

Extract the layer pixel values of the locations where the detection points of the first comparison group and the second comparison group are located respectively, and conduct numerical comparisons;

When the two detection points of the first comparison group have the same value and the two detection points of the second comparison group have the same value, it is determined that there is no texture between the first comparison group and the second comparison group. , no coloring processing is performed;

When the two detection points of the first comparison group have different values and the two detection points of the second comparison group have the same value, it is determined that there is texture between the first comparison group and the second comparison group does not There is texture; the window frame is divided into two according to the mid-perpendicular line of the first comparison group, and half of the window frame where the detection point with a value of 1 in the first comparison group is located is filled with color, and the other half is not perform coloring;

When the two detection points of the first comparison group have different values and the two detection points of the second comparison group have different values, it is determined that there is a common texture between the first comparison group and the second comparison group. The window frame is divided into two parts based on the mid-perpendicular line of any two detection points with different values among the four detection points of the first comparison group and the second comparison group, and the two values are 1. The half of the window frame where the detection point is located is filled with color, and the half where the two detection points with a value of 0 is is not filled with color;

Wherein, the color filling of the window frame is the texture information of the corresponding window frame.

As an improvement to the above solution, the texture information of all window frames in the window frame mask is spliced to obtain a sub-texture map of the corresponding layer, including:

After each window frame completes the color filling process of the area where it is located, the texture information of all window frames is merged according to its position in the window frame mask to form a sub-texture map of the layer where it is located.

As an improvement to the above solution, the weight superposition of all sub-texture maps to synthesize the total texture of the target image includes:

As an improvement of the above solution, the method also includes:

By setting the RGB value range, texture color blocks are filtered on the total texture map to obtain the final texture map output result.

In a second aspect, embodiments of the present invention provide an image texture extraction device, including:

The base conversion module is used to convert the color value of each pixel of the target image into a base to obtain a digital array of each pixel; among which, the base of the color value of the pixel after conversion is lower than the base of the color value of the pixel before conversion. system;

Layer decomposition module: used to decompose the target image into several layers according to the digital array;

A sub-texture map extraction module, used to extract texture information using a window frame mask for each of the layers, and obtain a sub-texture map of each of the layers;

The total texture extraction module is used to superimpose the weights of all the sub-texture maps to synthesize the total texture map of the target image.

In a third aspect, embodiments of the present invention provide an image texture extraction device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. The processor executes the The computer program is used to implement the image texture extraction method as described in any one of the first aspects.

In a fourth aspect, embodiments of the present invention provide a computer-readable storage medium. The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled. The image texture extraction method as described in any one of the first aspects is executed.

Compared with the existing technology, the beneficial effect of the embodiments of the present invention is that by performing base conversion on the color value of each pixel of the target image, a digital array of each pixel is obtained, and according to the digital array, the target image is converted into Decompose it into several layers; use a window frame mask to extract texture information for each layer to obtain a sub-texture map of each layer; superpose the weights of all the sub-texture maps and synthesize them The total texture map of the target image; the present invention increases the number of layers of decomposition of the target image by reducing the color value of the pixels of the target image, thereby effectively improving the completeness and accuracy of image texture extraction.

Description of the drawings

In order to explain the technical solution of the present invention more clearly, the drawings needed to be used in the implementation will be briefly introduced below. Obviously, the drawings in the following description are only some implementations of the present invention. For ordinary people in the art, For technical personnel, other drawings can also be obtained based on these drawings without exerting creative work.

Figure 1 is a flow chart of an image texture extraction method provided by an embodiment of the present invention;

Figure 2 is a schematic diagram of an exploded image provided by an embodiment of the present invention;

Figure 3 is a schematic diagram of a window frame mask provided by an embodiment of the present invention;

Figure 4 is a schematic diagram of the detection point of the window frame provided by the embodiment of the present invention;

Figure 5 is a schematic diagram of texture filling provided by an embodiment of the present invention;

Figure 6 is a schematic diagram of layer merging provided by an embodiment of the present invention;

Figure 7 is a schematic diagram of image texture information provided by an embodiment of the present invention;

Figure 8 is a schematic diagram of layer weight overlay provided by an embodiment of the present invention;

Figure 9 is a schematic diagram of an image texture extraction device provided by an embodiment of the present invention;

Figure 10 is a schematic structural diagram of an image texture extraction device provided by an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

Please refer to Figure 1. An embodiment of the present invention provides an image texture extraction method, which can be called the Multi-layer Grids Contralateral Opponent extraction method (MGCO), which specifically includes:

S1: Convert the color value of each pixel of the target image into a base to obtain a digital array for each pixel; where the base of the color value of the pixel after conversion is lower than the base of the color value of the pixel before conversion;

Texture is actually the difference in optical properties between two glossy surfaces. Texture will exist if and only if there is a color difference between the two glossy surfaces. Based on this, texture can be detected by detecting the color difference value between two light surfaces. Specifically, images can be represented in color using a variety of different color spaces, such as HSV, HSB, CMY, L*a*b*, etc. In the implementation of the present invention, RGB color space is used as an example to present the process steps of the present invention, but in principle, the method implemented in the present invention can be applied to any other color space to realize texture information extraction of images.

Further, the color value of each pixel of the target image is converted into a binary system to obtain a digital array, including:

In this embodiment of the present invention, the color value of each pixel of the target image can be converted from decimal to N, with N<10, thereby improving the layer separation of the image. For example, convert to binary, ternary, quaternary, etc. The smaller the selected base, the more layers the target image is decomposed, and the higher the accuracy of texture extraction.

Further, the color value of each pixel of the target image is converted into an N-ary system to obtain a digital array of each pixel, including:

Perform binary conversion on the color values of each pixel of the target image to obtain a 3×M-bit digital array for each pixel;

In the embodiment of the present invention, it is preferable to convert the color value of the decimal pixel into binary. Since the RGB color space consists of three color channels: red, green, and blue, the value of each color channel is an integer of [0, 255]. Value, a total of 256 color scale values. That is, each color channel can be represented by 8-bit binary. If all three color channels are represented in binary, the RGB value of each pixel can be expressed as a digital array composed of 3×8 total of 24 0/1 signals. Therefore, the color value comparison between two glossy surfaces can be understood in computer language as the numerical comparison between two sets of 3×8 digital arrays. The difference value between the two digital arrays is the texture value between the two glossy surfaces.

Embodiments of the present invention take advantage of the simplicity and clear contrast of binary signals to convert the task of identifying complete texture information into a task adapted to computer language, so that the computer can more simply and accurately extract all texture information in the target image.

For example, for a pixel with a color value of RGB (50, 120, 200), the binary number array corresponding to the color value is presented in the following table.

in:
R50＝128×0+64×0+32×1+16×1+4×0+2×1+1×0;
G120＝128×0+64×1+32×1+16×1+4×1+2×0+1×0;
B200＝128×1+64×1+32×0+16×0+4×1+2×0+1×0;

S2: Decompose the target image into several layers according to the digital array;

Furthermore, after binary conversion, according to the digital array of each pixel, the digits of each color channel in all pixels are clustered and merged into one layer, resulting in a total of 24 layers.

Among them, one bit of each color channel can generate a layer. The color value of each pixel is composed of three color channels: R, G, and B. Each color channel has an 8-bit 0/1 number. Name each number according to its "channel + binary digit". For example, the number located at the 6th digit of the R channel is named R6, the number located at the 4th digit of the G channel is named G4, and so on. Each pixel contains R1~R8, G1~G8, B1~B8, a total of 24 0/1 numbers. The digits of each color channel in all pixels are clustered and merged separately. For example, the R1 digital clustering of all pixels is combined into one layer, the R2 digital clustering of all pixels is combined to form a second layer, and so on. Finally, You will get 24 layers. Each layer has the same number of pixels as the original image, but the pixel values only have two values: 0 and 1.

Name each layer with the bits that make up the layer. For example, a layer composed of G5 digital clusters of each pixel is called a G5 layer, and so on. The 24 layers are layers R1~R8, G1~G8, and B1~B8, as shown in Figure 2.

S3: Use a window frame mask to extract texture information for each layer, and obtain a sub-texture map of each layer;

Further, weight values of all the sub-texture maps are superimposed to synthesize a total texture map of the target image.

Assign a weight to the sub-texture map: assign a weight to the sub-texture map according to the binary digit reflected in the name of the sub-texture map. The relationship between the weight y and its binary digit n is y=2^(n-1). For example: the weight of the G5 sub-texture map y(G5)=2^4=16. That is, the weight of the G5 sub-texture map is the green value 16.

All the sub-texture maps are superimposed according to their weights to form a total texture map of the target image.

For example, the total texture map is obtained by superimposing the weights of the sub-texture maps corresponding to the 24 layers, as shown in Figures 7 and 8. The total texture map obtained at this time will describe the texture of the target image, and intuitively use RGB values to express the color difference value of each texture point.

Taking Figure 4 as an example to illustrate the principle of texture weight superposition in the window frame. Among them, the RGB values on the left side of the window frame in Figure 8 are 200, 90, and 20, and the RGB values on the right side of the window frame are 210, 220, and 240. Then the sub-texture maps corresponding to the 24 layers are weighted according to their weights. After the values are superimposed, the total texture color difference between the left and right sides of the window frame can be obtained as:

R=layer R5+layer R4+layer R2=-16+4-2=-10;

G=layer G8+layer G3+layer G2=-128-4+2=-130;

B=layer B8+layer B7+layer B6+layer B3=-128-64-32+4=-220.

In the implementation of the present invention, based on the three channels of R, G, and B of the binary digital array with 8 digits each, the target image is divided into 24 layers, and then the texture is extracted for each layer, and then each image is The weights of the sub-texture maps of the layer are superimposed to obtain the total texture map, which can effectively improve the accuracy of image texture extraction and ensure that the extracted texture is complete and refined.

In an optional embodiment, S3: Use a window frame mask to extract texture information for each layer, and obtain a sub-texture map of each layer, including:

Among them, the point located on the boundary of the window frame is taken as the detection point to detect the pixel value of the layer at the location of the point, and the two detection points on opposite sides of the window frame are used as a pair of comparison groups.

Further, a first comparison group and a second comparison group are set at the boundary of each window frame, wherein the line connecting the two detection points of the first comparison group is connected to the line connecting the two detection points of the second comparison group. The angle between the lines is 360°/2n, where n represents the number of comparison groups;

In the embodiment of the present invention, the window frame mask (contralateral antagonistic window frame mask) is composed of several window frames of customized sizes. For example, in the embodiment of the invention, if the window frame is set to a size of 3×3 pixels, then for an original target image of 11×11 pixels, each layer is covered by a mask composed of 5×5 window frames. Then each window frame covers a total of 9 pixels of 3x3 of the image, as shown in Figure 3.

The structure of the window frame: As shown in Figure 4, each window frame has at least two comparison groups. The characteristic points of the comparison group are: 1) A comparison group consists of two detection points located on the boundary of the window frame. The points are symmetrical along the midpoint and located on opposite sides of the window frame; 2) The angle between the orientations of the two comparison groups is 360°/2n, where n represents the number of comparison groups. Taking the setting of the first comparison group A and the second comparison group B as an example, the texture information extraction of the window frame is explained:

Texture recognition: Extract the values of the pixels where the detection points of comparison groups A and B are located, and compare the values; if the two detection points of comparison groups A and B have the same value (both are 0, or both are 1), then It is determined that there is no texture between the comparison groups A and B; if the two detection points of the comparison group A or B have different values (one is 0 and the other is 1), it is determined that there is texture between the comparison group A or B;

Texture extraction: The window frame has a total of 4 detection points of comparison groups A and B to extract texture from the covered pixel area. According to the different detection results of comparison groups A and B, the following texture results are output:

1) If the two detection points of comparison group A have the same value, and the detection points of comparison group B have the same value: the window frame has no texture and no color filling.

2) If the detection points of comparison group A have different values and the detection points of comparison group B have the same value: the window frame is divided into two parts according to the mid-vertical line of comparison group A, and the half of the window frame where the detection point with a value of 1 in comparison group A is located Carry out coloring and leave the other half uncolored;

3) If the detection points of comparison group A have the same value and the detection points of comparison group B have different values: the window frame is divided into two according to the mid-vertical line of comparison group B, and the half of the window frame where the detection point with a value of 1 in comparison group B is located Carry out coloring and leave the other half uncolored;

4) If the detection points of comparison group A have different values and the detection points of comparison group B have different values: the window frame is divided into two according to the mid-perpendicular line of any two detection points with different values among the four detection points, and the two Half of the window frame where the 1 value detection point is located is filled with color, and the other half is not filled with color;

For example, assume that comparison group A is horizontally oriented, and comparison group B is vertically oriented.

When the detection points of comparison group A have the same value and the detection points of comparison group B also have the same value, it is determined that there is no texture in the window frame;

When the detection points of comparison group A have different values and the detection points of comparison group B have the same value, the window frame is divided into two parts by the mid-perpendicular line of comparison group A. At this time, if the left detection point of comparison group A is 1 and the right detection is 0, then the left half of the window frame will be filled with color and the right half will remain colorless; if the 1 value is at the right detection point, the right side of the window frame will Do color filling and keep the left side colorless;

When the detection points of comparison group A have the same value and the detection points of comparison group B have different values, the window frame will be divided into two parts with the mid-perpendicular line of comparison group B. At this time, if the upper detection point of group B is 1 and the lower detection point is 0, the upper half of the window frame will be filled with color and the lower half will remain colorless; otherwise, the lower half will be filled with color and the upper half will remain colorless. color;

When the detection points of comparison group A have different values and the detection points of comparison group B also have different values, the window frame is divided into two with the mid-perpendicular line of the two detection points that are different values among the four detection points, and the two detection points are divided into two parts. The half side where the 1 value detection point is located is filled with color. At this time, if the left detection point and the upper detection point are 1, the upper left side of the window frame will be filled with color, and the lower right side will remain colorless; if the right detection point and the lower detection point are 1, then the right side of the window frame will be filled with color. The lower side is filled with color, and the lower left side remains colorless;

Taking the midpoints of the four sides of the window frame as four detection points, and a pair of detection points on opposite sides as a comparison group, the four detection points of the window frame can be divided into vertical ones. A pair of detection points and a pair of detection points in the horizontal direction. Compare the values of the pixels where a pair of detection points are located. If any pair of detection points has outliers in the comparison, it means that there must be texture between the two detection points, that is, there is texture in the window frame; If there are no outliers in the comparison between the two pairs of detection points, there is no texture in the window frame. For example, the values of the pixels located at a pair of detection points in the vertical direction of a window frame on layer B1 are divided into 1 and 0, and the values of the pixels located at a pair of detection points in the horizontal direction are 0 and 0, then there is a texture in the window frame; For another example, the values of pixels located at a pair of detection points in the vertical direction of another window frame on layer B1 are divided into 0 and 0, and the values of pixels located at a pair of detection points in the horizontal direction are 0 and 0, then there are no pixels in the window frame. Texture is present. Specifically, the window frame can be divided into two parts using the mid-perpendicular line of a pair of detection points with different pixel values as the dividing line, and the half of the window frame where the pixel value is 1 is filled with color, and the pixel value in the window frame is 0. The half of the window frame is not filled with color; the window frame is filled with color according to the color weight of the corresponding layer. The color weight depends on the color channel corresponding to the layer and the weight corresponding to its number. Its weight y is the same as the color of the window frame. The relationship between the digit n is y=2^(n-1). For example, the window frame of layer R1 is filled with R=1, the window frame of layer G5 is filled with G=16, and the window frame of layer B8 is filled with B=128.

Considering that the texture within the window frame not only has color difference attributes, but also orientation attributes. In the implementation of the present invention, the angle between the two pairs is also set to 360°/2n, where n represents the number of detection points in the comparison group to help determine the orientation of the texture. At this time, the color-filling process on the window frame specifically includes:

Taking layer R1 as an example, set a horizontal and vertical comparison group for each window frame in layer R1. When only a pair of detection points are outliers, that is, one is 1 and the other is 0, then The mid-perpendicular line of a pair of outlier detection points is the orientation of the texture; when both pairs of detection points are outliers, the orientation of the texture is the diagonal line of the window frame. There are 16 combinations of two pairs of detection points. Both pairs of comparison groups have the same value, that is, when they are both 0 or 1, it means there is no texture in the window frame. Among them, the texture orientation within the window frame is shown in Figure 5. By setting multiple pairs of detection points to compare pixel values, the accuracy of the texture orientation within the window frame can be improved.

In an optional embodiment, the texture information of all window frames in the window frame mask is spliced to obtain a sub-texture map of the corresponding layer, including:

Among them, the sub-texture map is named according to the layer in which it is located. For example, the sub-texture map of the R1 layer is called the R1 sub-texture map, and so on. As shown in Figure 6.

In this embodiment of the present invention, the total texture map of the target image is the result of the superposition of texture phase weights in the window frames of 24 layers. Since there are 8 possibilities for the texture direction of the window frame on each layer, and the texture direction of the same window frame on different layers may be different, so after multiple layers are superimposed, the texture of the window frame in the total texture map It is often in the form of a rice grid. Suppose the area of the window frame is S, and the minimum color block unit of the total texture map is a right triangle with an area of S/8. Therefore, the present invention extracts the texture of the target image by using the window frame as the unit, and the obtained total texture map is The resolution will be 8 times higher than the window frame mask. When the window frame is set to 3x3 pixels, the area of each pixel is S/9, and the minimum resolvable color patch area of the total texture map is S/8. It can be inferred that the minimum color patch area of the total texture map is 1.125 times that of a single pixel. It can be seen that when the window frame is set to 3x3 pixels, the final texture image is close to the single-pixel level in detail resolution and has a detail resolution similar to the original image, making the extracted texture information better. completeness and precision.

Traditional edge detection algorithms are good at edge detection for grayscale images with obvious textures, but have poor edge detection effects for color images or natural environment images with weak or complex textures. In comparison, the present invention has excellent texture extraction performance for various image types, including abstract images, scatter plots, black and white images, color images, character images, environment images, etc.

Because the total texture map extracted by the present invention not only retains all texture information, but also uses RGB values to intuitively reflect the color difference characteristics of each texture point. Therefore, we can use the rule that the textures of the same object often have similar color difference properties to separate the textures of different objects in the target image (illustration required). It also has great application prospects in subsequent image processing technologies such as object separation and object recognition.

In an optional embodiment, the method further includes:

This method can flexibly filter the texture information of the total texture map by setting the RGB value range, thereby obtaining texture results with different levels of simplicity.

In the embodiment of the present invention, the total texture map contains all the texture information of the target image, and uses intuitive RGB values to display the color difference characteristics of each texture point, making the texture data contained in the total texture map highly mineable. space and extremely easy operability. For example, by setting RGB value filtering conditions in different numerical ranges, the total texture map is filtered through the RGB value filtering conditions, thereby obtaining different texture results. For example, set the RGB value filtering condition of R, G, B∈ (8, 255], that is, filter out all texture color patches with RGB values less than 8 at the same time; or further improve the RGB value filtering condition, set R, G, B∈ (16, 255] RGB value filtering conditions, filter out texture color blocks with RGB values less than 16 at the same time, and further streamline the texture. By setting different RGB value filtering conditions, the colors in the total texture map that do not meet the RGB value filtering conditions are Blocks are deleted directly, and the image formed by the remaining color blocks is the result image with clear texture direction, making the texture result have customized characteristics.

At the same time, since the total texture map in the embodiment of the present invention includes all texture information with color difference values 1-255, therefore Textures that are easily overlooked/difficult to recognize even with the naked eye can be detected.

In other embodiments, a sub-regional ranking filtering method can also be used to filter the total texture map, specifically including:

Divide the total texture map into i columns, j rows and ixj sub-regions, and perform color block screening independently on each sub-region. The filtering conditions are set to delete all color blocks with a color difference value ranking lower than n% in the sub-region, and filter out extremely weak textures with RGB values less than m. Among them, n and m are preset constants.

Filtering the total texture map through the sub-region ranking screening method can ensure that each sub-region has a certain texture retention, so that relatively strong textures in weak texture areas can be retained, and these relatively strong textures are often affected by light and shadow. Important textures that are difficult to extract due to environmental interference, thereby avoiding interference from ambient lighting on texture extraction.

Embodiment 2

Referring to Figure 9, an embodiment of the present invention provides an image texture extraction device, including:

The system conversion module 1 is used to convert the color value of each pixel of the target image into a system to obtain a digital array of each pixel; where the system of the color value of the pixel after conversion is lower than the system of the color value of the pixel before conversion. system;

Layer decomposition module 2: used to decompose the target image into several layers according to the digital array;

The sub-texture map extraction module 3 is used to extract texture information using a window frame mask for each of the layers, and obtain the sub-texture map of each of the layers;

The total texture extraction module 4 is used to superpose the weights of all the sub-texture maps to synthesize the total texture map of the target image.

In an optional embodiment, the system conversion module is used to convert the color values of each pixel of the target image into N systems to obtain a digital array for each pixel;

Among them, the color scale values of the R, G, and B color channels of each pixel are expressed in N-ary form respectively, and the RGB value of each pixel is expressed as a 3×M-bit digital array; where M represents the maximum size of the pixel. The corresponding digit of the color value in N base.

In an optional embodiment, the system conversion module includes

A binary conversion unit, used to perform binary conversion on the color values of each pixel of the target image to obtain a 3×8-bit digital array for each pixel;

The layer decomposition module includes:

The same digit clustering unit is used to cluster the digits of each color channel in all pixels into one layer according to the digital array of each pixel.

In an optional embodiment, the sub-texture map extraction module includes:

A layer texture extraction unit is used to cover each layer with the window frame mask; wherein the window frame mask is composed of several window frames;

The window frame texture extraction unit is used for each layer and each window frame to use at least two comparison groups to extract the texture of the layer area where each window frame is located, and obtain the texture information of the layer area where each window frame is located; wherein, each indivual The comparison group consists of two detection points on opposite sides of the window frame;

The window frame splicing unit is used to splice the texture information of all window frames in the window frame mask to obtain the sub-texture map of the corresponding layer.

In an optional embodiment, a first comparison group and a second comparison group are set at the boundary of each window frame, wherein a line connecting the two detection points of the first comparison group and the second comparison group are The angle between the lines connecting the two detection points in the group is 360°/2n, where n represents the number of comparison groups;

The window frame texture extraction unit includes:

A numerical comparison unit, used to respectively extract the layer pixel values of the locations where the detection points of the first comparison group and the second comparison group are located, and perform numerical comparison;

When the two detection points of the first comparison group have different values and the two detection points of the second comparison group have different values, it is determined that there is texture between the first comparison group and the second comparison group. texture;

In an optional embodiment, the window frame splicing unit is specifically configured to add the texture information of all window frames according to the window frame mask after each window frame completes the color filling process of its area. Merge at the middle position to form the sub-texture map of the layer where it is located.

In an optional embodiment, the total texture extraction module is configured to superpose weights of all the sub-texture maps to synthesize a total texture map of the target image.

In an optional embodiment, the device further includes:

The texture filtering module is used to filter the texture information of the total texture map according to the preset RGB value range to obtain the final total texture map.

It should be noted that the implementation principles and technical effects of the embodiments of the present invention are the same as those of the first embodiment, and will not be repeated here.

Embodiment 3

Referring to Figure 10, an embodiment of the present invention provides an image texture extraction device, including at least one processor 11, such as a CPU, at least one network interface 14 or other user interface 13, a memory 15, and at least one communication bus 12. The communication bus 12 is used to implement connection communication between these components. Among them, the user interface 13 may optionally include a USB interface, other standard interfaces, and wired interfaces. The network interface 14 may optionally include a Wi-Fi interface and other wireless interfaces. The memory 15 may include high-speed RAM memory, and may also include non-volatile memory (non-volatile memory), such as at least one disk memory. The memory 15 may optionally include at least one storage device located remotely from the aforementioned processor 11 .

In some embodiments, memory 15 stores the following elements, executable modules or data structures, or a subset thereof, or an extended set thereof:

Operating system 151, including various system programs, is used to implement various basic services and process hardware-based tasks;

Program 152.

Specifically, the processor 11 is configured to call the program 152 stored in the memory 15 to execute the image texture extraction method described in the above embodiment, such as step S1 shown in FIG. 1 . Alternatively, when the processor executes the computer program, it implements the functions of each module/unit in each of the above device embodiments, such as a binary conversion module.

Exemplarily, the computer program may be divided into one or more modules/units, and the one or more modules/units are stored in the memory and executed by the processor to complete the present invention. The one or more modules/units may be a series of computer program instruction segments capable of completing specific functions. The instruction segments are used to describe the execution process of the computer program in the image texture extraction device.

The image texture extraction device may be a computing device such as VCU, ECU, BMS, etc. The image texture extraction device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the schematic diagram is only an example of the image texture extraction device and does not constitute a limitation on the image texture extraction device. It may include more or fewer components than shown, or combine certain components, or Different parts.

The so-called processor 11 can be a microprocessor (Microcontroller Unit, MCU) central processing unit (Central Processing Unit, CPU), or other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), or application-specific integrated circuits. (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general processor can be a microprocessor or the processor can be any conventional processor, etc. The processor 11 is the control center of the image texture extraction device and uses various interfaces and lines to connect the entire image texture extraction device. various parts of.

The memory 15 may be used to store the computer programs and/or modules. The processor 11 implements the above by running or executing the computer programs and/or modules stored in the memory and calling the data stored in the memory. Describes various functions of image texture extraction equipment. The memory 15 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the storage data area may store Store data created based on the use of the mobile phone (such as audio data, phone book, etc.), etc. In addition, the memory 15 may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart memory card (Smart Media Card, SMC), secure digital (Secure Digital, SD) Card, Flash Card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

Wherein, if the integrated modules/units of the image texture extraction device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above embodiment methods, and can also implement A computer program instructs relevant hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above method embodiments. Wherein, the computer program includes computer program code, which may be in the form of source code, object code, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory) , Random Access Memory (RAM, Random Access Memory), electrical carrier signals, telecommunications signals, and software distribution media, etc.

Embodiment 4

Embodiments of the present invention provide a computer-readable storage medium. The computer-readable storage medium includes a stored computer program. When the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the first step. The image texture extraction method described in any one of the embodiments.

It should be noted that the device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physically separate. The unit can be located in one place, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between modules indicates that there are communication connections between them, which can be specifically implemented as one or more communication buses or signal lines. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

The above is the preferred embodiment of the present invention. It should be pointed out that for those of ordinary skill in the art, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also regarded as It is the protection scope of the present invention.

Claims

An image texture extraction method, characterized by including:

Convert the color value of each pixel of the target image into a base to obtain a digital array for each pixel; where the base of the color value of the pixel after conversion is lower than the base of the color value of the pixel before conversion;

Decompose the target image into several layers according to the digital array;

Use a window frame mask to extract texture information for each layer, and obtain a sub-texture map for each layer;

All the sub-texture maps are superimposed by weights to synthesize a total texture map of the target image.
The image texture extraction method according to claim 1, characterized in that the color value of each pixel of the target image is converted into a binary system to obtain a digital array, including:

Convert the color values of each pixel of the target image into N-ary systems to obtain a digital array for each pixel;

Among them, the color scale values of the R, G, and B color channels of each pixel are expressed in N-base form, and the RGB value of each pixel is expressed as a 3×M-bit digital array; among them, the N-base is smaller than the color value Current forwarding, M represents the number corresponding to the maximum level value of the color channel in the N system.
The image texture extraction method according to claim 2, characterized in that the color value of each pixel of the target image is converted into an N system to obtain a digital array of each pixel, including:

Perform base conversion on the color values of each pixel of the target image to obtain a 3×M-bit digital array for each pixel;

Then, according to the digital array, the target image is decomposed into several layers, including:

According to the digital array of each pixel, the digits of each color channel in all pixels are clustered and merged into one layer.
The image texture extraction method according to claim 3, characterized in that, using a window frame mask to extract texture information for each layer, and obtaining a sub-texture map of each layer, including:

The window frame mask is used to cover each layer; wherein the window frame mask is composed of several window frame units;

Each window frame uses at least two contrast groups to extract the texture of the layer area where it is located, and obtain the texture information of the layer area where each window frame is located; among them, each contrast group consists of two opposite sides of the window frame. Detection point composition;

The texture information of all window frames in the window frame mask is spliced to obtain a sub-texture map of the corresponding layer.
The image contour detection method according to claim 4, characterized in that a first comparison group and a second comparison group are set at the boundary of each window frame, wherein the line connecting the two detection points of the first comparison group The angle between the line connecting the two detection points in the second comparison group is 360°/2n, where n represents the number of comparison groups;

Then, each window frame uses at least two pairs of comparison groups to extract the texture of the layer area where it is located, and obtain the texture information of the layer area where each window frame is located, including:

Extract the layer pixel values of the locations where the detection points of the first comparison group and the second comparison group are located respectively, and conduct numerical comparisons;

When the two detection points of the first comparison group have the same value and the two detection points of the second comparison group have the same value, it is determined that there is no texture between the first comparison group and the second comparison group. , the window frame is not filled with color;

When the two detection points of the first comparison group have different values and the two detection points of the second comparison group have the same value, it is determined that there is texture between the first comparison group and the second comparison group does not There is texture; the window frame is divided into two according to the mid-perpendicular line of the first comparison group, and half of the window frame where the detection point with a value of 1 in the first comparison group is located is filled with color, and the other half is not perform coloring;

When the two detection points of the first comparison group have different values, and the two detection points of the second comparison group have different values, Then it is determined that there is a common texture between the first comparison group and the second comparison group, and the window frame is determined by any two of the four detection points of the first comparison group and the second comparison group that are mutually different values. The middle vertical line is used as the boundary, and the window frame is divided into two parts. The half of the window frame where the two detection points with a value of 1 are located is filled with color, and the half of the window frame where the two detection points with a value of 0 are located is not filled with color.

Wherein, the color filling result of the window frame is the texture information of the corresponding window frame.
The image texture extraction method according to claim 5, wherein the texture information of all window frames in the window frame mask is spliced to obtain a sub-texture map of the corresponding layer, including:

After each window frame completes the color filling process of the area where it is located, the texture information of all window frames is merged according to its position in the window frame mask to form a sub-texture map of the layer where it is located.
The image texture extraction method according to claim 1, characterized in that, superposing weights of all sub-texture maps to synthesize a total texture map of the target image includes:

All the sub-texture maps are superimposed by weights to synthesize a total texture map of the target image.
The image texture extraction method according to claim 1, further comprising:

By setting the RGB value range, texture color blocks are filtered on the total texture map to obtain the final texture map output result.
An image texture extraction device, characterized by including:

The system conversion module is used to convert the color value of each pixel of the target image into a system to obtain a digital array of each pixel; among which, the system of the color value of the pixel after conversion is lower than the system of the color value of the pixel before conversion. ;

Layer decomposition module: used to decompose the target image into several layers according to the digital array;

A sub-texture map extraction module, used to extract texture information using a window frame mask for each of the layers, and obtain a sub-texture map of each of the layers;

The total texture extraction module is used to superpose the weights of all the sub-texture maps to synthesize the target map. The overall texture map of the image.
An image texture extraction device, characterized in that it includes: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the following is implemented: The image texture extraction method according to any one of claims 1-9.
A computer-readable storage medium, characterized in that the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute as claimed in claim 1- The image texture extraction method described in any one of 9.