FI20216300A1 - Method and arrangement for determining the proportion of bark in wood chips - Google Patents

Abstract

This description describes a method and an arrangement implementing the method for determining the proportion of bark in wood chips. The method comprises the steps in which the chip pieces are spread on the conveyor so that the chip pieces are separated from each other and the upper sides of the chip pieces traveling on the conveyor are photographed with the first camera to detect the shell on the upper side. Chips falling from the end of the conveyor are filmed from their underside with a second camera in order to reveal the bark on the underside of the chips, and the image information produced by the first and second cameras is combined into chip-specific identification information so that, based on the identification information, any bark on the top and bottom of the chips is identified for each chip. The total amount of bark in the wood chips per chip piece is estimated based on this identification information. (Figure 2a)

Description

Method and arrangement for determining the proportion of bark in wood chips

Field of the invention — The invention is related to monitoring the quality of wood chips, and in particular to determining the proportion of bark in the chips.

Background

In the production and utilization of wood chips, large cost savings can be achieved by optimizing the quality of the wood chips. The amount of bark in the wood chips can be seen as a measure of quality that can be optimized: on the one hand, an excessive amount of bark in the wood chips weakens the quality of the pulp produced from the wood chips, and on the other hand, completely removing all the bark from the wood chips also increases the loss of wood material. Thus, one may want to be able to monitor and, if necessary, change the amount of bark in the chip very precisely.

Chips traveling on the conveyor can be photographed with a camera, and by analyzing the image information produced by the camera, the proportion of the amount of shell in the 5 chips can be estimated. As a challenge in this kind of image information based

However, in N's analysis, there are many changing parameters associated with the chips traveling on the conveyor, such as the size, shape, wood quality,

N rin, and change of position. Consequently, part of the chip remains unrecognized/classified

E: lemma or is often misclassified, in which case the exact value of shell size

S of the number of women in the woodworking industry is often difficult to produce.

O

S

Quick description

This description presents the method and arrangement intended for determining the amount of shell, which solve the challenges presented above. The method and arrangement according to the invention is characterized by what is presented in the existing patent claims.

In the method and the arrangement that implements it, the wood chips traveling on the conveyor are imaged from both sides (from the top and later from the bottom when it falls from the conveyor belt) in order to determine the portion of the shell in the wood chips. When the wood chips are photographed with two cameras, the measurement accuracy improves, because the material contained on both sides of the wood chips can be found out.

The measurement accuracy can be further improved by combining the image information captured by the cameras in such a way that the opposite sides of the chip captured by the cameras can be identified in the method as the same chip. Each piece of wood chips thus forms a model, which contains the information needed to identify the shell — for both sides of the piece of wood chips in question. On the basis of such chip-specific information, a very accurate value can be calculated for the amount of bark in the chip.

N Image list

S

N 20 — In the following, the invention is explained in detail by referring to the accompanying drawings, in which: &

E Figures 1a — 1c show diagrams of example chip pieces and

Figures 2a and 2b show the separate application form of the system implementing the method according to this description.

OF

Detailed description

This description describes the features of the method and arrangement according to the invention in more detail.

One key feature of the method and arrangement is that the amount of bark is examined per chip piece, i.e. each chip piece can be evaluated separately and the amount of bark contained in each chip piece can be given its own value. For example, a quantity can be represented by a value representing surface area, volume, mass or another measurement quantity. At its simplest, the amount can be represented by a value that can be chosen between two options (where the options can represent the meanings "includes shell" and "does not contain shell).

Definitions related to the application

The context of the method and arrangement and the vocabulary used in the description are opened next with the help of figures 1a — 1c.

Figures 1a - 1c show simplified isometric examples of wood chips. Figure 1a shows three mutually perpendicular dimensions for the chip piece 10: length £, width W and thickness 7. The chip piece usually has two surfaces opposite each other, extending in the plane defined by the width direction W and the length direction L. Such surfaces define two normally rectangular, planar "sides" for the chip. The log piece therefore has a pack-

S 20 —in the mouth direction 7 two opposite sides. Planar shape of a chip

Therefore, it is practically always placed on the conveyor in such a way that one of these sides o is towards the conveyor plane of the conveyor and the other is away from the conveyor plane. Unless

In other words, the term "chip side" (or simply "side") refers to

S in this description to the surface as specified above. In Figure 1a, one of these 2 25 — sides is visible and is represented by the reference number 10a. In addition, in Fig. la

N shows one of the two opposite ends 10b of the chip piece and one of the chip piece

N from two opposite sides 10c.

The chip piece 10 in Figure 1a is entirely wood. On the sides of the chip piece made of wood, a parallel grain pattern 12 resulting from the fibers in the direction of the wood's longitudinal growth can typically be observed. The grain pattern 12 is parallel to the length £ of the chip piece. In some application forms of the method, root patterns can be used as an aid in determining the longitudinal direction /. The ha- Kepala 10 of the pattern is made entirely of wood and the grain pattern 12 can be seen on both sides of the chip piece 10.

The longitudinal ends 10b of the wood chips typically form widthwise, planar surfaces. The planes of the ends 10b are typically at right angles to the planes of the sides 10c, but are slanted so that the sides 10c are in the form of diagonal trapezoids. The contour of the chip piece (I. silhouette) described directly in the thickness direction, defined by the ends 10b and the sides 10c, is therefore typically rectangular.

Although Figure 1a shows a piece of wood chips where the length L is greater than the width

W the shape of wood chips is not limited to such pieces. The width W of the chip can also be greater than the length /. However, the thickness of the chip 7 is typically smaller than both the length Z and the width W.

The dimensions described above and the properties and definitions associated with them also apply to figures 1b and 1c. Any subsequent references to such — dimension dimensions and definitions are also intended to be understood above

N as described. a = Figure 1b shows a piece of wood chips with dimensions corresponding to the description in Figure 1a, part of which, however, consists of the shell. Bark refers to the layer that surrounds and protects the wood of the tree, such as the bark. In Figure 1b only

S 25 part of the thickness of the chip piece 7 is the bark. The upper side of the hakapala is wood 14, when

And the lower side on the opposite side is made of beetle 16. The upper side and

The lower half are only relative designations for the opposite sides of the search field to make it easier to understand the pictures. The upper half refers to the part seen on top in Figure 1b, and the lower half to the part below the upper half. When on the conveyor, the top side of the chip can be seen. If the amount of bark was monitored with one camera, the bark of such a piece of wood chips could go undetected if the piece was on the conveyor with the bark side down. Figure 1c, on the other hand, shows a piece of wood chips, where part of the piece is also a shell. Unlike Figure 1b, however, in Figure 1c, the bark material 16 is on the side of the chip and only extends to part of the width of the chip

W. The rest of the chip is wood14. On the other hand, the bark material 16 extends in figure 1c to the thickness 7 of the entire chip so that the bark is visible on both sides of the chip. Although the examples in Figures 1a — 1c do not show it separately, the chip piece can also consist entirely of the shell.

Describing log fires

In order to determine the proportion of bark in the wood chips, the method according to this description and the equipment implementing the method can be used, where the pieces of wood chips are described from both sides. The method includes the steps, — in which the chip pieces are spread on the conveyor in such a way that the chip pieces settle essentially in one layer. In other words, the chip pieces are not on top of each other but are separate from each other. The individual pieces of wood chips are therefore separable from each other. The upper sides of the wood chips traveling on the conveyor are photographed from the upper side with the first camera to detect the bark on the upper side and from the lower side with the second camera to detect the bark on the lower sides of the wood chips. The conveyor can be formed in the method and arrangement, for example

S conveyor device that can be part of a longer conveyor arrangement. The conveyor device can pre-

N to be a conveyor belt with which the wood chips pass on one side (top side)

Q is shot with the first camera. The conveyor device may be arranged to terminate

E 25 — in such a way that the pieces of wood chips fall from its end into the collection container or into the next

Fuck it. Another camera can be adapted to record the first

O the opposite side (bottom side) of the wood chips to the side the camera is shooting

O when falling from the end of the conveyor device. However, another type of conveyor can also be used, as long as the chip pieces can be spread on it in such a way that the chip pieces can be photographed from both sides and are distinguishable from each other in the image information.

Figures 2a and 2b show diagrams of one application form of the arrangement that implements the method. Figure 2a shows a schematic view directly from the side, where the arrangement comprises a conveyor device 20 in the form of a conveyor belt, which has a conveyor platform on which wood chips can be spread into loose wood chips 21. The arrangement has a first camera 22 arranged to film the upper sides of the wood chips advancing on the conveyor, a second camera 23 arranged to film the conveyor 20 the undersides of the chips 21 falling from the end. The top and bottom sides of the wood chips are illuminated in Figure 2a with lamps 24 and 25. A partial enlargement of Figure 2a shows a piece of wood chips similar to Figure 1b, where the top side is wood 14 and the bottom side is bark 16. Figure 2b shows a simplified diagram of the objects being transported on the conveyor 20, shown directly from above, from each other from the separate pieces of wood chips 21. Cameras 22 and 23 or lights 24 and 25 are not shown in figure 2b —. In the wood chips of Figure 2b, the grain patterns of the wood parts of the wood chips 21 are also not shown.

Identification of bark in a piece of wood chips

Identification of the bark of wood chips based on image information can be implemented in several ways. In the method, the image information produced by the cameras is analyzed in order to find and separate individual pieces of wood chips.

From N. For this purpose, the analysis aims to separate the chips from their background.

N At the same time, the contour (I. silhouette) is determined for each piece of wood chips. The separation between the chip and the background can also be based on, for example, the chip and the background

I to a clear shadow and/or contrast difference. Conveyor belt (in figure 2a and 2b belt

E 25 20) can for example be blue in color. In this way, a contrast can be created with the pieces of ash, whose wood and bark parts stand out well against such a background,

N especially when the lighting is chosen appropriately.

OF

In addition to separating the wood chips from their background, the material on the sides of the wood chips is analyzed based on the image information produced by the cameras. The image recognition algorithm (or algorithms) used can, for example, be adapted to simply classify the described surface of the wood chip into one of two predefined classes: bark or wood. Alternatively, the classification determined by the identification data can be more multi-level (only wood, partly bark, full bark) and/or the classification may involve other parameters, such as numerical values related to the dimensions and quality of the wood chips. In shell detection, for example, already known image recognition algorithms can be used. Identification can be done, for example, by differentiation based on the difference in color and/or contrast between bark and wood.

Cameras that photograph wood chips from both sides (cameras 22 and 23 in the example of Figure 2a) can be, for example, line or matrix cameras that photograph a color or black-and-white image. Cameras can photograph wood chips in the visible light range and/or outside it (such as in the ultraviolet or infrared range). The lighting (illuminators 24 and 25 in the example of Figure 2a) is preferably of high intensity in order to achieve a high contrast. In addition, the wavelengths of the light produced by the lighting can be chosen so that, on the one hand, the pieces of wood chips differ from their background (such as the transport plane of the conveyor) and on the other hand, the bark and wood parts of the pieces of wood chips differ from each other in terms of color and/or contrast.

Chips can be illuminated with, for example, Xenon, LED, halogen or UV lighting

For N sim. The transport platform of the conveyor (such as the conveyor belt) can, for example, be

Make it blue in color, so that the pieces of wood chips stand out well from the conveyor plane, especially if they are illuminated with, for example, blue or yellow light.

OF

E 25 — Although figure 2a shows the individual, point-like light sources lighting the wood chips

For example, wood chips can also be lit with flat light sources, such as

With an O lighting matrix. In this way, possible shadows and reflections can be

Minimize these disturbances in image information.

In some cases, the general color shade or contrast of wood and bark parts can be close to each other, and differentiation based solely on color and/or contrast is not enough.

For example, the wood may have darkened as a result of moisture damage. In this case, separating wood and bark parts can be more challenging.

As already mentioned earlier, the wood sections have a structure formed by wood fibers in the direction of the tree's longitudinal growth, which appears as a parallel grain pattern visible on the surface of the wood chips. Bark parts do not have such a root pattern, so when separating wood and bark parts, the method can also utilize the detection of a root pattern. To detect the existence of a root pattern on the surface of a chip, for example, a machine vision algorithm can be used, which has been taught to recognize patterns consisting of parallel parallel lines similar to the root pattern.

Depending on the form of application, the separation based on the color and/or contrast difference and the separation based on the detection of the root pattern can be used together or — separately to identify the surface material of the chip. Particularly good results can be obtained by combining the above-mentioned separation methods, i.e. separation based on color/contrast difference and separation based on the detection of a root pattern.

Modeling the log

When the chip pieces and the objects on their opposite sides (bark and/or wood) have been identified from the image streams, the proportion of bark in the chip can be determined. simple

N at most from the image information produced by the two cameras = it would be possible to separately count the chips containing the shell and the two obtained

It would be possible to add up N result values to get the final result. In this = approach, however, pieces of wood chips showing bark on both sides

S 25 — purple, would be included in the result value of both cameras. Thus, such

O chip pieces will be counted in the final result twice, which ai-

O would throw an error in the final result value.

In order to avoid this error, the image information produced by the first and second cameras is combined in the method into chip-specific identification information.

Combining can be based, for example, on the location information of detected wood chips. In addition, the contour information of the wood chips can be used for merging.

Based on the identification information, it is possible to detect the bark that may be contained on both sides of the chip pieces (i.e. the top and bottom). This identification information can be seen as a digital modeling of the chip, which contains the parameter values necessary to identify the features to be measured. In other words, the identification data presents a model of each chip piece, which is based at least on image information from both sides of the chip piece. Depending on the form of application, this model can be enriched with other information.

Based on the identification information or the model, the number of pieces of wood chips containing bark can be accurately calculated. Since the image information provided by the two cameras is combined individually for each chip piece, information on the content of both sides of each chip piece is available with the help of identification information. Depending on the form of application, the information can be a simple classification between two options (contains shell / does not contain shell) or it can be a numeric value representing the amount of shell. The numerical value can, for example, be the surface area of the shell deduced from the image information of the camera. — With the help of identification information, it is possible to distinguish pieces of wood chips that are completely wood from pieces of wood chips that have bark on one side. Calculated in this way, the hakepa-

S lats that have a shell on both the upper and lower sides are only counted together

N times, when the total number of pieces of wood chips containing bark in the wood

The measurement accuracy of the Q quantity estimate improves. The total amount can be pre-

E 25 — as a sign as a ratio in relation to all wood chips or only in relation to it

S for wood chips. In applications that are produced in a single

O numerical value for the amount of bark in the chip, this numerical value can be used

O for calculating the size of the shell in the chip.

The calculation steps of the method described above can, for example, be done with the calculation unit belonging to the system (not shown in figures 2a and 2b). The calculation unit can be adapted to receive the image information produced by the first and second cameras and to combine the image information into identification information so that based on the identification information, the bark that may be contained on the top and bottom sides of the chip pieces is identified for each chip piece.

Utilization of thickness information

In the method and its application forms described above, the thicknesses of the wood chips transported by the conveyor can also be measured. The thickness information can be combined as part of the identification information/model. When the thickness of the chip piece is known, the proportion of bark in the piece can be estimated more precisely. The amount of bark contained in wood chips can be calculated on the basis of the completed identification information for each chip piece, and the total amount of bark in the wood chips can be calculated based on the estimates of the amount of bark per piece of wood chips. By combining the measured (or — estimated or assumed in another way) shell thickness with the surface area of the shell detected by the camera, an estimate of the amount of shell in the chip can be calculated, for example in the form of the volume (or mass) of the shell in the piece. Similarly, the amount of wood contained in each piece of wood chips can be calculated. The calculated proportions of bark and wood can be used to calculate a more accurate overall estimate of the proportion of bark in wood chips. To the extent that the volume or mass cannot be directly calculated based on the available information, one can rely on selected basic assumptions

Here's the structure of the chip. If the shell does not, for example, reflect the images produced by the cameras

Based on N information, extend to the entire thickness of the chip from one side to the other,

Q, for example, it can be assumed that there is a certain percentage of bark in the piece, such as

E 25 — for example, 50% of the thickness of the chip. 5 One way to measure the thickness of wood chips is to form wood chips with a conveyor

N with a laser installed at an oblique angle to the angle of view of the imaging camera

N pattern (e.g. a line parallel to the width of the conveyor belt). In Figure 2a, the laser 26 creates such a pattern for the conveyor 20. In Figure 2b, the pattern 26a produced by the laser is shown. When there is no chip on the conveyor near the laser beam, a straight line is formed on the conveyor. On the other hand, when the conveyor has a piece of wood chips near the laser, the line is shaped according to the cross section of the piece of wood chips, and based on this shape, the thickness of the piece of wood chips in question can be estimated. In some — application forms, the first camera can be adapted to photograph both the color/contrast of the wood chips and the thickness-measuring line produced by the line laser. Alternatively, the line produced by the laser can be detected by its own camera or other measuring arrangement. For example, thickness can be measured in other ways, such as lidar.

Other uses

The method described above and the equipment that implements it are suitable not only for measuring the amount of shell but also for other things. The method and equipment can, for example, be used to detect deviations in wood chips. In this way, you can get a lot of additional information about the saw that produces the wood chips itself and the condition of the saw blades.

When the silhouette of the wood chips (and possibly also the grain pattern) is known, the shape of the wood chips can also be analyzed. The method can, for example, identify contours of wood chips that deviate from the rectangular shape, such as a semi- = 20 — trapezoidal contour (i.e. the ends of the wood chips are not

N parallel). Trapezoidal pieces of wood chips, for example, indicate = that there is too much wood in the felling from time to time and that the uppermost trees

N pods can move/turn laterally. This kind of information can be used, for example, in controlling the capacity of the peeling line. The method

S 25 — part of it can therefore be an artificial intelligence algorithm that has been taught to recognize semi-

O neck-shaped contours of wood chips. Information about the contour shape of the chip

O can be added as part of the identification information / template of the chip piece.

Blunt damage and dullness of the blades can also be identified based on the image information produced by the cameras. Burst damage refers to damage caused by compression applied to the chip, which is manifested as the curling of the ends of the chip. Similarly, the dullness of the blades of a chip-producing chip can be evaluated based on the appearance of the cutting surface at the end of the chip. The method can therefore comprise an artificial intelligence algorithm that has been taught to detect visual signs related to these fault situations from the image information produced by the cameras.

OF

O

OF

OF

O

OF

I

=

O

O

0

O

OF

O

OF

Claims (7)

Patent Claims 1. A method for determining the proportion of bark in wood chips, which method includes the steps in which: - the chip pieces are spread on the conveyor so that the chip pieces are separated from each other, - the upper sides of the chip pieces traveling on the conveyor are photographed with the first camera to detect the bark on the upper side, - the chip pieces falling from the end of the conveyor are photographed from below with a second camera in order to detect the bark on the underside of the wood chips, - the image information produced by the first and second cameras is combined into piece-specific identification information so that the possible bark on the top and bottom of the pieces of wood chips is identified on the basis of the identification information, and - the total amount of bark in the wood chips is estimated based on the piece-specific identification information . 2. The method according to claim 1, where the evaluation of the total amount of bark comprises steps in which - the number of pieces of wood chips containing bark is calculated for each wood chip based on the identification information, and N - the total amount of bark is estimated in the wood based on the estimated number of pieces of wood chips containing bark. ; O = S N 25 3. The method according to claim 1 or 2, where: - — the thicknesses of the wood chips transported by the conveyor are measured, - the thickness information is combined into part of the identification information, - the amount of bark contained in the chip pieces is estimated for each chip piece based on the identification information, and - the total amount of bark is estimated in the chip based on the estimates of the amount of bark per chip piece. 4. The method according to one of claims 1 to 3, where the merging of identification data comprises steps in which - the image information produced by the first and second cameras related to both sides of the chip piece is analyzed in order to separate the chip pieces from their backgrounds and from each other, - the image information produced by the first and second cameras is analyzed in order to determine the amount of bark contained on both sides of the wood chips, and - the analyzed amount of bark contained on both sides of the wood chips is combined into chip-specific identification information. 5. The method according to one of claims 1 to 4, where: N - the image information is analyzed to find root patterns on the opposite sides of the pieces O N 20, N T - the analysis of the root pattern is used to deduce the number of shells O N contained in the sides of the chip piece. T O O 0 O N O N 6. Equipment for determining the proportion of bark in the wood chips, which equipment comprises: a) a conveyor device with a conveyor platform on which the wood chips can be spread into loose chips b) a first camera arranged to photograph the upper side of the chips moving along the conveyor, c) a second camera arranged to photograph the chips falling from the end of the conveyor bottom side, and d) when the computing unit is adapted to receive the image information produced by the first and second cameras, where the computing unit is adapted - — to combine the image information produced by the first and second cameras into chip piece-specific identification information so that, based on the identification information, the top and bottom sides of the chip pieces are identified on a per chip piece basis bark that may be contained, and - to estimate the total amount of bark in the chip based on the chip piece-specific identification information. 7. An arrangement according to claim 6, where the arrangement further comprises N N e) a lighting arrangement which is adapted to illuminate the wood chips in the camera imaging in order to improve the separation of the wood chips and their background. I = O O 0 2 25 N O N