KR101118516B1 - Method to predict strength of species-mixed wood using radiation technique - Google Patents

Abstract

The present invention relates to a method for predicting the strength of mixed species wood using radiation, and to classify the knots and healthy parts by imaging the wood with radiation, and to classify the knots through the connection evaluation of knots, and to select a representative cross section for each knot, By converting the cross section into equivalent densities, the strength of all species can be accurately predicted without knowledge of the species. According to the present invention can be used to predict the strength of the wood of any species, regardless of species, it can be used for structural purposes.

Wood, equivalent density, density conversion, knots, strength

Description

Method to predict strength of species-mixed wood using radiation technique

The present invention relates to a method for predicting the strength of all kinds of wood without distinguishing the species by analyzing the image taken using radiation and expressing the equivalent density.

The domestic timber industry, which relies on imports for most of its timber supply, needs to automate the wood processing process in order to increase the yield and worker operation efficiency as raw material prices continue to rise.

Recognition of defects with the naked eye of a skilled worker, which is a method currently used, may change the decision criteria according to the psychological state of the worker or the degree of fatigue, resulting in a change in the consistency of work speed and quality. Huber et al. (1985) measured the recognition rate of lumber surface defects with the naked eye for workers in a furniture factory, and measured 59 ~ 74% according to the skill of the workers.

And, unlike the visual method, when the defect was recognized using machine vision, the sound material discrimination rate was 99%, and the defect discrimination accuracy was 88% (McMillin, 1984).

However, when the defects on the surface of the wood are recognized using the naked eye or machine vision, the location of the defects can be accurately determined.However, the strength of the wood cannot be predicted because it is difficult to determine the size of the defects up to the interior of the wood. Difficult disadvantages arise.

Of course, the strength of the wood can be predicted by calculating the knotty penetration ratio as in the previously proposed method of predicting the strength of wood using the X-ray scanner (Application No. 10-2009-0022347). There is a need for prior information such as regression equations, and a database of knot density for each species.

In the case of wood, since the performance of structural materials cannot be controlled by changing the manufacturing method such as the ratio of raw materials and processing methods, such as steel and concrete, grade classification is performed to classify wood having similar performance.

In order to minimize the variation, the classification rules are set and classification is made for each production area, and the design values are calculated for each species and grade.

However, this method is available only when a certain species is produced in large quantities in a specific area, and in fact, various species grow together in the forest, and thus there is a disadvantage such as classification of species.

Therefore, there is a need for a method capable of predicting the strength of any species by predicting the strength of wood without prior knowledge of the species and classification of species.

In order to solve the problems of the prior art, the present invention provides a method capable of predicting the strength of all species of wood without distinguishing the species of wood by acquiring images of wood using radiation and quantifying the obtained images with equivalent density. There is a purpose.

In order to achieve the above object, in the method of predicting the strength of all kinds of woods of the present invention, a two-dimensional image of wood is obtained by using radiation, and the image is separated into an image of a knot and a sound image (part of the knot). The separated images can be analyzed to quantify each section with equivalent density to predict the strength of wood.

The present invention relates to a method for estimating the strength of all species of wood without distinguishing the species, comprising the steps of detecting wood knots by radiating the wood with radiation, dividing the wood by pixels, and evaluating the pixel connectivity of the knots. And separating the knot images, and selecting the representative cross sections for each knot, and calculating a cross-sectional reduction ratio for each pixel of the wood including the representative cross sections for each knot. . And converting a cross-sectional reduction ratio of the wood including the representative cross section into a density, calculating a ratio of the converted density to a reference density to form a modified cross section, and then, based on the modified cross section, in the longitudinal direction of wood. And calculating the same density and predicting the strength of the wood.

In the present invention, the pixel connectivity evaluation may classify each knot image by identifying the knot image from the wood. That is, the knots can be distinguished by assigning identifiers such as numbers and letters, respectively, according to the number of knots.

In the present invention, the step of selecting the representative cross section for each knot may be obtained by the following equation, and the representative cross section for each knot may select a representative cross section for each knot classified by an identifier by pixel connectivity evaluation. In addition, the representative cross section may be formed by a row including the most pixels in one knot, and the knots in which the representative cross section is selected may be represented by the sound portion except for the representative cross section.

(Where, ρ _{i, j} : ρ ^* _i (p): the density of the (i, j) position pixel of the pth knot determined by the pixel connectivity evaluation, h _{i, j} : the pth knot determined by the pixel connectivity evaluation) (i, j) The distance from the center of the timber width of the pixel at position -h _{i, j} = | (m / 2) -i | (m is the number of pixels corresponding to the timber width)-, ρ _c : Sound density average, i: row number of the radiographic image (width direction, 1 ~ m), j: column number of the radiographic image (length direction, 1 ~ n))

In the present invention, the step of calculating the cross-sectional reduction ratio for each pixel of the wood including the representative cross section for each knot is given by the following formula in consideration of knots existing within 140 mm to 160 mm in the longitudinal direction of the wood about the representative cross section. After obtaining the respective cross-sectional reduction ratios, the cross-sectional reduction ratios for the entire wood can be obtained by adding each cross-sectional reduction ratio. At this time, the knot around the representative cross section can also be formed as the representative cross section.

I, row number of the radiographic image (width direction, 1 to m), j: column number of the radiographic image (length direction, 1 to n), p: knot number, r _{i, j} (p): row i Cross-section reduction ratio by column k in column j, x _p : distance between column j and column k in mm, ρ ^* _i (p): i-th acting density of representative section of k-th array, ρ _{c is the} mean density of sound areas of the analyzed wood)

In the present invention, the step of converting the cross-sectional reduction ratio into a density may refer to the following equation. At this time, the healthy portion that does not contain knots can be said to have a high density and thus high strength. However, since the portion containing the knot is high in density but low in strength, the cross-section reduction ratio of the knot is determined to determine the density of the healthy portion and the knot. By converting the density into equivalent density, the strength of the wood can be predicted using the equivalent density of the wood including sound areas and knots.

(I: row number (1 to m) of the radiographic image, j: column number (1 to n) of the radiographic image, R _{i, j} : cross-sectional reduction ratio due to knots of the position (i, j), ρ _c : Average density of sound areas of analytical wood, ρ _{i, j} : Density at position (i, j), ρ _{i, j} `: Converted density at position (i, j))

In the present invention, the step of calculating the equivalent density of the deformed cross section may refer to the following equation, and the deformed cross section is formed by the density of the sound portion and the converted knot density (representing a low density of the portion detected as the knot). It may be configured as a 3D image based on the 2D image.

: j열의 등가밀도, ρ_basis : 기준밀도)(T _j : cross section secondary moment (deformation) of deformed section for column j or area (compression, tensile) of deformed section, I _gross : section secondary moment (deformation) or cross sectional area (compression, tensile) for wood ),

: j equivalent column, ρ _basis : reference density)

In the present invention, the strength of the wood may include any one selected from bending strength, tension, and compression. Through the equation of calculating the equivalent density, it is possible to predict the bending strength, and by substituting the area for the equation instead of the second moment, the tensile and compression of the wood can be calculated. That is, the strength prediction method of the present invention can be made irrespective of the species of wood, and can predict the strength of tensile and compression using not only the flexural strength but also the area.

According to the present invention, by predicting the strength of the wood through the deformation cross-section including the converted density of knots and the density of the healthy portion, there is an effect that can predict the strength of the wood without classification.

Preferred embodiments of the present invention will be described with reference to the accompanying drawings. In adding reference numerals to components of the following drawings, it is determined that the same components have the same reference numerals as much as possible even if displayed on different drawings, and it is determined that they may unnecessarily obscure the subject matter of the present invention. Detailed descriptions of well-known functions and configurations will be omitted.

1 is a view showing an apparatus for predicting the bending strength of wood according to an embodiment of the present invention.

Referring to FIG. 1, in order to measure the wood 100 as a two-dimensional image, the image may be measured while penetrating through the radiation detector 110 and the radiation generator 120.

At this time, the radiation emitted from the radiation generator 120 includes all radiation, such as x-rays, gamma rays, wood 100 is a radiation detector 110 and the radiation generator 120 in a thin direction of the wood as shown in the figure Can penetrate through.

Hereinafter, how to predict the strength of wood using the apparatus of FIG. 1 will be described below.

2 is a flowchart illustrating a method of predicting bending strength of wood according to an embodiment of the present invention.

Referring to FIG. 2, in the step S200 of capturing an image, a 1D image of wood is photographed using radiation such as gamma ray and X-ray, and the photographed 1D image is continuously connected to the 2D image of wood. Implement the original image.

In the step of detecting knots (S210), the knots are detected using image analysis on the original image of the wood photographed in step S200.

For details, refer to FIGS. 3A to 3C below.

In the separating of the sound field image (S212), only the image of the portion except for the knot detected in step S210 is obtained, and the average density value of the sound portion is obtained by converting each pixel value of the image composed of the sound portion into a density value.

Separating the knot image (S220) corresponds to step S212 and separates only the image of the knot detected in step S210 to impose an identifier for each knot. In this case, the identifier may be arbitrarily selected by the user, such as numbers, letters, or English.

The step S230 of selecting a representative cross section for each knot is to select a representative cross section for each of the knots separated by identifiers. Refer to to select the column with the largest I _j value as the representative cross section.

That is, when defining images of one column ((1 * m) unit) as one cross-section according to Equation 1 below, I _j for the cross-section position j is calculated for the cross-sections in which p knots exist. A cross section having a large value is selected as the representative cross section of the p-th knot (see FIG. 4B).

Ρ ^* _i (p): density of the (i, j) position pixel of the pth knot determined by the pixel connectivity evaluation, h _{i, j} : (i, j) of the pth knot determined by the pixel connectivity evaluation -H _{i, j} = | (m / 2) -i | (m is the number of pixels corresponding to the timber width)-, ρ _{c is the} mean density of the sound region of the analytical timber, i: row number of the radiographic image (width direction, 1 ~ m), j: column number of the radiographic image (length direction, 1 ~ n))

The knot pixels included in other columns except one column set as the representative cross section for one knot are represented by 0 and the representative cross sections are represented by the corresponding density to form the representative cross section kinematic density array.

For details of the representative cross section selection, refer to FIGS. 4A and 4B.

Computing the reduction ratio (S240) calculates the reduction ratio in order to consider the knots adjacent to each other for the wood image including the representative cross-section selected in step S230.

The knot density is about twice that of the healthy part, and since the knot can be assumed to be a cross section defect, it can be seen that the high density in the knot density array causes a high cross-sectional reduction.

In addition, it is necessary to add the value of the knot density array using adjacent knot information because several knots may exist in the wood at narrow intervals and additional strength reduction may occur due to adjacent knots.

In consideration of the characteristics of the knot, the effect of p knot on a specific position (i, j) is calculated as in Equation 2 below.

: p번 옹이의 대표단면의 i번째 행의 밀도, ρ_c : 분석 목재의 건전부 평균 밀도)(I: row number of the radiographic image (width direction, 1 to m), j: column number of the radiographic image (length direction, 1 to n), p: knot number, r _{i , j} (p): row i Cross-section reduction ratio due to pth knot of j column pixels, x _p : Distance between jth k pth knots in mm,

: density of the i-th row of the representative cross section of knots, ρ _c : mean density of sound areas of the analyzed wood)

In Equation 2 above, the length of the stringing knot is considered to be less than 140 mm to 160 mm in the length of the wood, and more precisely, the knot less than 150 mm should be considered. Therefore, only knots existing within 75 mm of the longitudinal direction of the knot to be measured are considered.

Then, by adding the effect of each knot on the (i, j) position cell, the cross-sectional reduction ratio Ri _{, j} can be finally obtained as shown in Equation 2 below.

(R _{i, j} : Cross-section reduction ratio due to the knots of the i-row j-column pixels, r _{i, j} (p): Cross-section reduction ratio due to the p-th knot of the i-row j-column pixels)

In the step of converting the density (S250) in consideration of the average density value of the healthy portion calculated in step S212 and the cross-sectional reduction ratio of the knots calculated in step S240 can be obtained the density of the wood using the following Equation 4.

Where i is the row number of the radiographic image (width direction, 1 to m), j is the column number of the radiographic image (length direction, 1 to n), and R _{i , j} is due to the knots of the position (i, j). Cross-sectional reduction ratio, ρ _c : mean density of sound areas of the analyzed wood, ρ _{i, j} : density at (i, j), ρ _{i, j} `: converted density at (i, j))

Referring to Equation 4 above, the pixels detected as knots are converted to have a low density when the cross-sectional reduction is large, and to be equal to a healthy density when the cross-sectional reduction is small, considering the cross-sectional reduction.

Calculating the equivalent density (S260) may calculate the density ratio through the following equation 5 based on the density converted in step S250, and multiply the calculated density ratio by the thickness of the wood to implement the deformation cross-section (Equation 6 Reference).

n _{i , j} = ρ _{i, j} `/ ρ _basis

(i: row number of the radiographic image (width direction, 1 ~ m), j: column number of the radiographic image (length direction, 1 ~ n), ρ _{i, j} `: converted density at position (i, j), ρ _basis : reference density, n _{i , j} : density ratio of (i, j) position)

In the above equation 5, the reference density is 0.5 g / cm 3, which is eliminated by multiplying the reference density in subsequent calculations, and thus is not affected by the magnitude of the reference density.

w` _{i, j} = w × n _{i, j}

(i: row number of the radiographic image (width direction, 1 ~ m), j: column number of the radiographic image (length direction, 1 ~ n), w` _{i , j} : of the deformation cross section with respect to position (i, j) Thickness ratio, w: thickness of wood)

Hereinafter, a method of calculating the equivalent density may be described in detail with reference to FIGS. 5A to 5B.

Searching for the minimum equivalent density (S270), the step of searching for the smallest equivalent density of wood according to the deformation section implemented in step S260 to predict the strength of the wood (S280). In other words, the smallest equivalent density shows the smallest strength among the wood, which can be judged to be the most likely to be broken when the bending load occurs.

Therefore, the density of the healthy part is used as it is to express wood as density, and the density of knot is considered to have a decrease in cross section, so that the higher the density, the higher the strength, and the lower the density, the lower the density. It can be predicted the bending strength of wood through having.

3A to 3C are diagrams illustrating knotty connectivity evaluation according to an embodiment of the present invention.

Referring to FIG. 3A, an original image 100 of wood using radiation is shown. In this case, when the wood is photographed using radiation, the higher the density, the darker the color, and the lower the density, the lighter the color.

Usually, the knot density is about 2 times higher than the density of the healthy part where no knot is present, so the knotty part having a high density appears dark and the healthy part appears bright.

Accordingly, since the original image of the wood photographed by the radiation appears as a dark knot 300 in the bright sound portion, it is possible to separate the knot and sound image separately.

Referring to FIG. 3B, only the knot 300 is separated from the original image 100 of wood, and a plurality of knots 300 are included in one image.

Referring to FIG. 3C, after separating the image of the knot in FIG. 3B, a plurality of knots are distinguished by imposing an identifier for each knot. The identifier used at this time may be arbitrarily imposed by the user such as numbers, letters, and figures.

4A to 4B are diagrams showing a method for setting a representative cross section of the knot according to an embodiment of the present invention.

Referring to FIG. 4A, the entire image of wood including knot images separated by identifiers is illustrated in pixels in FIG. 3C. In the present invention, numbers are used as identifiers, but the present invention is not limited thereto.

5A to 5D are views showing a method for constructing a deformed cross section according to an embodiment of the present invention, and FIG. 5A is a view showing a shape of knots occupying a cross section in a wood original image photographed by radiation. 5b is the density measured in the cross section. (This example is a case where there is no adjacent knot in the longitudinal direction of the wood for the sake of easy deformation cross section configuration.)

Referring to FIG. 5B, it can be seen that the density of the portion in which the knot is present is higher than the density of the healthy part because the section reduction ratio is not considered.

Referring to FIG. 5C, after the representative cross section of the knot is selected, the cross-section reduction ratio of the knot is calculated and converted into density.

Referring to FIG. 5D, a deformation ratio may be configured by obtaining a density ratio based on the density converted in FIG. 5C and multiplying the thickness of wood (Equations 5 and 6).

5A to 5B illustrate a specific cross section, that is, one column, when the wood is divided into pixels, and the cross section is formed along the length of the wood by forming a cross section of all cross sections from column 1 to column n. Can be.

And, based on the deformation cross section, the equivalent density for column j can be calculated as in Equation 7 above, and the same calculation can be repeated from column 1 to column n to implement the equivalent density for the entire wood.

Equation 7 is the case of bending

, 및

, And

For compression, tension

,

의 두 식에 의해 계산될 수 있다. 이때, (T_j : j열에 대한 변형단면의 단면 2차 모멘트(휨) 또는 변형단면의 단면적, I_gross : 목재단면치수에 대한 단면 2차 모멘트(휨) 또는 단면적(압축, 인장), m : j열의 마지막 픽셀의 i행 번호, s : x선 영상의 1픽셀의 한 변의 길이(mm), h : 목재의 너비(mm), h_i : 목재의 중립축과 (i,j)셀까지의 거리(mm))를 나타낸다.

,

It can be calculated by two equations. At this time, (T _j : cross-sectional secondary moment (deformation) of the deformation cross section for row j or cross-sectional area of the deformation cross section, I _gross : cross-section secondary moment (bending) or cross-sectional area (compression, tension) for wood cross section dimension, m: i row number of the last pixel of column j, s: length of one side of 1 pixel of x-ray image (mm), h: width of wood (mm), h _i : neutral axis of wood and distance to (i, j) cell (mm)).

6a to 6b is a view showing the wood and the equivalent density thereof, including knots according to an embodiment of the present invention, Figure 6a is a view showing the wood, Figure 6b is a longitudinal direction of the wood (1 column ~ n columns) Is a graph showing equivalent density for each section by calculating the equivalent density with).

Referring to FIG. 6B, the smallest equivalent density correlates most closely with the strength characteristics of the wood, and the strength of the wood can be predicted through regression analysis.

7a and 7b is a graph comparing the prediction method of the present invention and the existing prediction method, Figure 7a uses the existing prediction method to know what species, and Figure 7b is applied to all species when the species is unknown The prediction method of the present invention can be used.

Referring to FIG. 7A, the flexural strength is predicted in the presence of species information, while R ² shows an accuracy of about 0.58, while FIG. 7B is also used to predict the flexural strength without information on the species for both larch and pine species. Nevertheless, R ² shows an accuracy of about 0.73.

As described above, it has been described with reference to a preferred embodiment of the present invention, but those skilled in the art various modifications and changes of the present invention without departing from the spirit and scope of the invention described in the claims below I can understand that you can.

1 is a view showing an apparatus for predicting the bending strength of wood according to an embodiment of the present invention.

2 is a flowchart showing a method of predicting the bending strength of wood according to an embodiment of the present invention.

3A to 3C are diagrams illustrating knotty connectivity evaluation according to an embodiment of the present invention.

4A and 4B illustrate a method of setting up a representative cross section of a knot according to one embodiment of the invention.

5A-5D illustrate a method for constructing a modified cross section in accordance with one embodiment of the present invention.

6a and 6b is a view showing the wood and the equivalent density thereof including a knot according to an embodiment of the present invention.

7A and 7B are graphs comparing the conventional prediction method with the prediction method of the present invention.

             <Explanation of symbols for the main parts of the drawings>

100: wood 110: radiation detector

120: radiation generator 300: knot

Claims (7)

Photographing wood with radiation to detect knots of the wood; Dividing the wood into pixels, and separating each knot image through pixel connectivity evaluation of the knots; Selecting the separated cross-section by each knot, and calculating a cross-sectional reduction ratio for each pixel of the wood including the representative cross-section of the knot by considering adjacent knots; Converting the reduction ratio of the cross section of wood including the representative cross section into a density; And Calculating a ratio of the converted density to a reference density to form a deformation cross section, calculating an equivalent density in the longitudinal direction of wood based on the deformation cross section, and predicting the strength of the wood; Strength prediction method of mixed species wood using radiation comprising a. The method of claim 1, wherein the pixel connectivity evaluation distinguishes knot images from the wood and identifies each of the knot images. The method of claim 1, wherein the step of selecting the representative cross section for each knot includes referring to the following equation.

(Where ρ ^* _i (p): the density of the (i, j) position pixel of the pth knot as determined by the pixel connectivity evaluation). h _{i, j} : The distance from the center of the wood width of the (i, j) position pixel of the pth knot as determined by the pixel connectivity evaluation. h _{i, j} = | (m / 2) -i | (m is the number of pixels corresponding to the width of the wood) ρ _c : sound density average of the analyzed wood. i: Row number of the radiographic image (width direction, 1 ~ m) j: column number of the radiographic image (length direction, 1 ~ n)) The method of claim 1, wherein the calculating of the cross-sectional reduction ratio for each pixel of the wood including the representative cross-section by each knot includes: Considering knots existing within 140 mm to 160 mm in the longitudinal direction of the wood based on the representative cross-section, the strength reduction method of the mixed species wood using radiation, characterized in that the addition of the cross section reduction ratio for each knot is obtained by the following equation.

(Where i is the row number of the radiographic image (width direction, 1 to m),      j: column number of the radiographic image (length direction, 1 to n),      p: knot number, r _{i, j} (p): cross sectional reduction ratio by the pth knot of the i row j column pixels, x _p : distance between column j and p knot (mm), ρ ^* _i (p): i-th acting density of the representative section of k p ρ _c : mean density of sound areas of the analyzed wood) The method of claim 1, wherein the step of converting the reduction ratio into density refers to the following equation.

(Where i is the row number of the radiographic image (1 to m)      j: the column number of the radiographic image (1 to n) R _{i, j} : Cross-section reduction ratio by knot of (i, j) position ρ _c : sound density average of the analytical wood ρ _{i, j} : density at position (i, j) ρ _{i, j} `: converted density at position (i, j)) The method of claim 1, wherein the step of calculating the equivalent density of the deformed cross section refers to the following equation.

(Where T _j : cross-sectional secondary moment (deformation) of the deformed section for column j or area of the deformed section (compression, tension), I _gross : Cross section secondary moment (bending) or cross section (compression, tensile) for wood,

: equivalent density of column j, ρ _basis : reference density) The method of claim 1, wherein the strength of the wood is any one selected from bending strength, tension, and compression.

Images