JP7766951B2 - Laminated wood inspection and sorting equipment - Google Patents

Description

The present invention relates to the technical field of wood laminate manufacturing.

In recent years, the number of cedar, cypress, and other trees planted in Japan's forests has grown, resulting in an abundant stock of timber in the country. Coupled with the movement toward a decarbonized society, there are high hopes for the use of timber as a building material.
When used as building materials, etc., it is required to inspect the quality and manufacture and supply products that meet the appropriate standards.
Wood is used in two forms: solid wood and laminated wood, such as plywood and glued lumber. Plywood is manufactured by cutting veneers from raw logs, drying them, gluing the veneers, stacking them, and pressing them. Plywood is made using different types of veneers, such as top, back, core, and slats. For example, top veneers have clean surfaces with few defects, while slats are thicker than the top veneer and may contain knots or discoloration. Furthermore, plywood is generally laminated with the grain direction perpendicular to the board. Therefore, odd-numbered layers, such as the top and back, run longitudinally, while even-numbered layers, such as slats, run perpendicularly to the board. These veneers are evaluated and sorted according to their intended use while still in their veneer state. The present applicant also proposed a method and apparatus for evaluating defects, discoloration, etc., using transmitted and reflected light and sorting them accordingly, as shown in Patent Document 1 (Patent Publication No. 6017390).

For laminated materials such as plywood, the number of layers and adhesive layers means that the irradiated light cannot penetrate, so inspection methods using transmitted light cannot be used and the evaluation methods described above cannot be applied.

Patent No. 6017390

The purpose of this invention is to develop an inspection device for laminated materials such as plywood made from wood.

1. A laminated wood lumber inspection device characterized by comprising a surface inspection device that uses vertical and oblique illumination light irradiated onto the surface of the laminated wood lumber, and an end grain inspection device that uses end grain illumination light irradiated onto the end grain of the laminated wood lumber.
2. The inspection device for laminated wood lumber according to 1., characterized in that the surface inspection device is equipped with inspection devices for both the front and back surfaces.
3. The surface inspection device is equipped with a vertical illumination means for illuminating the surface of the laminated wood lumber from a substantially vertical direction, an oblique illumination means for illuminating the surface of the laminated wood lumber from an oblique direction, and a surface imaging means,
The surface inspection device for laminated wood lumber described in 1. is characterized in that it separates a vertical light image and an oblique light image from image data captured by the surface imaging device, compares the two images of a common location on the laminated wood lumber, and determines the properties of the surface.
4. The end grain inspection device includes a long end grain inspection device equipped with a long end grain illumination and a long end grain imaging means, and a short end grain inspection device equipped with a short end grain illumination and a short end grain imaging means,
The long-length end grain inspection device determines the properties of the long-length end grain from the long-length end grain image,
1. The inspection device for laminated wood lumber according to 1., characterized in that the short-length end grain inspection device determines the properties of the short-length end grain from an image of the short-length end grain.
5. The laminated wood lumber inspection device according to 4., wherein the image for inspecting long or short end grains is a corrected long or short end grain image obtained by rearranging and correcting images taken from the long or short end grain imaging means in the thickness direction of the laminated wood lumber.
6. The laminated wood lumber inspection device according to 5., characterized in that if there is a portion in the corrected long or short end grain image where the thickness is different compared to the front and back, a second corrected long or short end grain image is used as the image for inspection, which is corrected by comparing it with the reference long or short end grain image to make it common with the layer of that portion.
7. The surface inspection device judges the surface based on information obtained by synchronizing the vertical light image, the oblique light image, the long end grain inspection image, and the short end grain inspection image.
The long-length end grain inspection device determines the long-length end grain based on information obtained by synchronizing the long-length end grain image with the image of the surface inspection device,
The short-length end grain inspection device determines the short-length end grain based on information obtained by synchronizing the short-length end grain image with the image of the surface inspection device,
7. The inspection device for laminated wood lumber according to any one of 4. to 6., wherein the comprehensive judgment is made based on a surface judgment, a long piece end grain judgment, and a short piece end grain judgment.
8. The inspection device for laminated wood lumber according to any one of 1. to 6., wherein the laminated wood lumber is any one of plywood, LVL, LVB, CLT, and thick plywood.
9. An inspection and sorting device for laminated wood lumber, comprising the inspection device for laminated wood lumber described in any one of 1. to 6. and a stacking device that separates and stacks the lumber based on the inspection results.
10. An inspection and sorting device for laminated wood lumber, comprising the inspection device for laminated wood lumber described in 7. and a stacking device for sorting and stacking the lumber based on the inspection results.
11. An inspection and sorting device for laminated wood lumber, which is arranged in this order: a long-length end grain inspection device, a surface inspection device, a short-length end grain inspection device, and a stacking device, each of which is continuously connected by a conveying device, and which is provided with an overall controller for controlling these devices;
The long end grain inspection device includes a conveying device that conveys wood laminated lumber vertically, a lighting device and an imaging device provided on the side of the conveying device, a material detector provided above the conveying device, and an encoder provided on the conveying device.
The surface inspection device includes a conveying device that conveys the wood laminated lumber laterally, a lighting device and an imaging device provided above and below the conveying device, a material detector provided above the conveying device, and an encoder provided on the conveying device.
The illumination device is provided with two types of illumination: a reflected light illumination device that reflects light incident almost perpendicularly to the imaging device, and an oblique light illumination device that illuminates the imaging device obliquely.
The short-length end grain inspection device is provided with a conveying device that conveys the wood laminated lumber laterally, and a lighting device and an imaging device are provided on the side of the conveying device, and the conveying device is conveyed synchronously with the conveying device of the surface inspection device,
The fractionation device has a plurality of deposition devices, each of which is successively provided by a transport device;
This is an inspection and sorting device for laminated wood lumber, characterized in that the comprehensive controller evaluates the laminated wood lumber based on three inspections: a long-length end grain inspection device, a surface inspection device, and a short-length end grain inspection device, and stacks the laminated wood lumber in each stacking device of the sorting device according to the evaluation.
12. The long end grain inspection device further includes a means for rearranging the captured images in the thickness direction of the wood laminated lumber to form a corrected long end grain image,
12. The inspection and sorting device for laminated wood lumber according to claim 11, further comprising a means for forming a corrected image of the short end grain by rearranging the captured images in the thickness direction of the laminated wood lumber.
13. The laminated wood lumber inspection and sorting device according to 12., characterized in that if there is a portion in the corrected long end grain image or the corrected short end grain image where the thickness is different compared to the front and back, a second corrected long end grain image or a second corrected short end grain image that has been corrected by comparing it with the reference long end grain image or the reference short end grain image so that it is common to the layer of that portion is used as the image for inspection.
In this invention, six sides of the laminated wood lumber are inspected and judged: the front and back, both sides of the long end grain, and both sides of the short end grain. Although it is described as an inspection of one side, such as "surface inspection," the "back side inspection" is basically the same, so an explanation of the other side will be omitted. Therefore, the front side includes the front and back sides, the surface inspection device includes an inspection device for the front side and an inspection device for the back side, and the long end grain and short end grain include both end grains.

1. The present invention provides an inspection device for laminated wood lumber that can evaluate the surface properties and end grain properties of laminated wood lumber using images captured by irradiating light and capturing reflected light.
By comparing images obtained using light reflected vertically and light reflected obliquely on the surface of laminated wood lumber, it is possible to compare the shadow conditions caused by unevenness and determine whether there are any live knots or knotholes.
2. Knots (knot holes), cracks, overlaps, etc. can occur at the end grain of laminated wood lumber, so these defects can be evaluated based on images of the reflected light from the end grain.
3. With the present invention, laminated veneer plywood can be inspected, evaluated, classified, and provided as thick plywood with an increased number of laminated veneers, multi-layer plywood, plywood, LVL, CLT, LVB, and laminated lumber, etc. This allows it to be provided as a base material, exterior material, or other material suited to its intended use.
4. The overall image of long and short end grains, which is created by capturing and synthesizing the individual divided images obtained from the line sensor while the laminated wood lumber is being transported, can be affected by the warping of the laminated wood lumber and the vibration of the transport conveyor, resulting in a curved, non-linear image. In such cases, a linear inspection image (primary end grain corrected image) can be generated by correcting the individual images by rearranging them in the thickness direction using the surface or median of the individual images as a reference. Because laminated wood lumber is a flat board, it can be restored to a linear state by rearranging them flat.
Furthermore, if there is a defect in the laminated wood lumber, it will appear as a recess in the primary end grain corrected image, but it may be difficult to determine which layer the defect is in. In such cases, a second correction is performed to generate a secondary end grain corrected image.
The second correction involves rearranging the individual images of the recesses to match the layers of the reference end grain image of the laminated wood lumber, thereby obtaining a secondary corrected image and clarifying the layer of the missing part. The reference end grain image can be generated by compressing the primary image correction, or can be based on the layers before and after the recess.
For example, if a primary end grain correction image is formed by focusing on the surface lines and arranging the surfaces of individual images from a line sensor in a straight line, a missing portion will occur on the lower side, but by comparing it with the reference end grain image and rearranging it so that the layers other than the missing portion match, the missing layer will become a recess.
Alternatively, if the primary cut edge correction image is created by focusing on the median of each individual image, the missing part will appear as a depression at the top or bottom. Even in this case, by comparing it with the reference edge image, the missing layer will become a depression by rearranging the layers other than the missing part so that they match.
Inspection accuracy is improved by using the primary end grain corrected image and then the secondary end grain corrected image.
5. Japanese coniferous trees such as cedar, pine, cypress, and larch, which were planted after the war, have grown and are increasing in volume as lumber. However, due to low lumber prices and insufficient care such as pruning, there is discoloration due to knots and oil, and the quality varies widely. By appropriately evaluating these trees using the present invention, it is possible to promote the use of these trees.

FIG. 1 is a diagram showing a plan view of a laminated wood lumber inspection and sorting device. FIG. 2 is a front view showing an outline of a laminated wood lumber inspection device. A diagram showing an inspection device for long end grain. A diagram showing an inspection device for the front and back surfaces and short end grain. 10A and 10B show examples of image generation obtained from imaging of front or back inspection. 1A and 1B are diagrams showing a schematic diagram of overlapping and example photographs. A diagram showing a schematic diagram of a tunnel and example photos. FIG. FIG. 10 is a diagram showing an example of a distorted image. Illustrative corrections: (a) correction for aligning individual images with the surface line; (b) correction for aligning individual images with the center line. 10A and 10B are diagrams illustrating two-stage correction using a compressed image as a reference image. 10A and 10B are diagrams showing two-stage correction based on the front and rear of a recess; 1 is a diagram showing an example of a laminated wood material.

This invention is a laminated wood lumber inspection device that can evaluate the surface and end grain properties of laminated wood lumber using images captured by irradiating light and capturing reflected light. By comparing images obtained using light reflected from a vertical direction and light reflected from an oblique direction, the surface of the laminated wood lumber can be evaluated to determine the presence of live knots, knotholes, etc. by comparing the state of shadows caused by unevenness.
Knot holes, cracks, overlaps, etc. can occur at the end grain of laminated wood lumber, so these defects are evaluated based on images of the end grain.
The present invention targets laminated wood lumber made by laminating wood boards, such as plywood made by laminating veneers, thick plywood made by increasing the number of laminated veneers, LVL, and CLT. It is possible to inspect and evaluate laminated wood lumber and provide classified products. This allows for the provision of products suitable for various uses, such as underlayment and exterior materials.
Japanese coniferous trees such as cedar, pine, cypress, and larch vary greatly in quality due to discoloration caused by knots and oil, but the present invention makes it possible to properly evaluate these trees and promote their use.

The present invention is an inspection device for laminated wood lumber that includes a surface inspection device that uses surface irradiation light irradiated onto the surface of the laminated wood lumber and an end grain inspection device that uses end grain irradiation light irradiated onto the end grain of the laminated wood lumber. In this invention, each inspection is performed while the laminated wood lumber is being transported, and the laminated wood lumber is managed so that its position can be identified by each inspection. For example, by managing each inspection device using a common time, the inspection positions of the laminated wood lumber can be synchronized and it can be evaluated as a single piece of laminated wood lumber.
The surface inspection device is equipped with a vertical illumination means that illuminates the surface of the laminated wood lumber from a nearly vertical direction, an oblique illumination means that illuminates the surface of the laminated wood lumber from an oblique angle, and a surface imaging means. Vertical light images and oblique light images are separated from continuously captured image data, and the two images of a common location on the laminated wood lumber are compared to determine the properties of that location. Since the shadows created by the two types of light are different for holes and irregularities, it is possible to determine whether the surface is flat or uneven. Furthermore, by using color images, it is also possible to determine rot, discoloration, etc.
The surface inspection device can be provided on both the front and back surfaces, and inspection devices with the same configuration can be provided on both the front and back surfaces.

The end grain inspection device includes a long end grain inspection device equipped with a long end grain illumination and a long end grain imaging means, and a short end grain inspection device equipped with a short end grain illumination and a short end grain imaging means.
The long-length end grain inspection device determines the properties of long-length end grains from images of long-length end grains. The short-length end grain inspection device determines the properties of short-length end grains from images of short-length end grains. Furthermore, the end grain can be determined by taking images of the front and back surfaces together.

The laminated wood lumber passes through inspection devices continuously as it is transported. During transport, the laminated wood lumber is detected by a detection sensor, and the position where the lumber passes through each inspection device is recorded by an encoder or the like. By aligning each inspection device according to time management, defects in specific parts of a single piece of laminated wood lumber can be identified and detected, and the entire lumber can be inspected.
By synchronizing the inspection signals of each inspection device with the signals of other inspection devices, it becomes possible to correlate and evaluate from multiple angles. For example, the overlap of the end grain and the bulge on the surface are synchronized, so the area of the overlap can be grasped.

When images are captured with a line sensor from warped laminated wood or when the wood is subjected to vibrations from a conveying device, even if the individual images captured in small increments are lined up, unevenness will appear and the image will be distorted, making it difficult to obtain a straight overall image and making the image difficult to evaluate (see Figure 9).
These distorted end grain images are corrected by rearranging the individual images in a straight line so that the end grain surface is straight. The corrected image of the long or short end grain (see Figure 10) (primary end grain corrected image) can be used as the evaluation image for highly accurate evaluation.
To correct the image to a straight line, you can focus on the top surface (surface) of each image and align them in a straight line. Alternatively, you can align them by focusing on the center. Since laminated wood lumber basically has a uniform thickness, you can obtain a straight, corrected image by aligning each image based on the surface, bottom, center, etc.

If there is a defect in a laminated wood lumber (especially in the top or bottom layer), it will appear as a dent in the linearly corrected primary end grain corrected image, but in order to identify which layer has the defect, the image is re-corrected to obtain a secondary end grain corrected image.
There are several methods for re-correction. For example, individual images obtained from a line sensor are rearranged to match a reference image, and a secondary end grain correction image is created, with the missing layer being treated as a recess. Another method focuses on the top and bottom layers, and rearranges the images to match the top or bottom layer that matches the layer before and after the recess. Yet another method is to use the layers before and after the recess as the reference layers, and rearrange the images to best match the layer structure of the recess.
Defects that cause depressions on the surface are often found in the top or bottom layers of laminated wood lumber. Defects in the middle layers often appear as tunnels.
Furthermore, when laminae or other plate materials are stacked, there is no risk of any loss of part of the lamina, so there is little need for secondary correction.

An overall judgment is made by combining evaluations of the long ends on the left and right, the short ends on the front and back, and the front and back surfaces.
The overall evaluation is determined by taking into consideration the type, number, and extent of defects, such as the number of knots and the extent of tunnels. The evaluation criteria may be changed depending on the purpose and tree species.

These evaluations and judgments can be made automatically through AI learning.
The initial criteria for judgment may be, for example, to prepare a large number of thick plywood boards, and for each board, images are taken and the end grain, front and back surfaces are input according to the evaluation items, and the system is trained.
By accumulating learning in accordance with actual operations, the accuracy of evaluation can be improved.

In the present invention, two surfaces, the front and back, and four end grain surfaces of the laminated wood lumber are inspected and evaluated.
By inspecting the front and back surfaces, it is possible to detect and evaluate the dimensions of laminated wood lumber, live knots, dead knots, holes, cracks, insect holes, rot, blade marks, and poor putty.
Inspection of the end grain allows for the detection and evaluation of overlaps, tunnels, and chips in the materials that make up the layers (such as veneers).
Examples of test items are shown in Table 1.
Laminated wood materials are materials made by laminating layers of wood such as wood veneers, and include plywood, thick plywood, multi-layer plywood, LVL, CLT, LVB, and laminated lumber.

Figure 1 shows a plan view of the laminated wood lumber inspection and sorting device SS.
The inspection and separation device SS includes an inspection device A and a separation and deposition device B. The inspection device A and the separation and deposition device B are arranged in series and are controlled by a comprehensive judgment controller 8.
The wood laminate inspection and sorting device SS is a device that continuously transports wood laminate 9, inspects and evaluates the wood laminate 9 during transport, and sorts and stacks the wood laminate 9 based on the evaluation results.

The inspection device A includes a long wood end grain inspection device 30, a front and back inspection device 40, and a short wood end grain inspection device 60.
The sorting and stacking device B is equipped with a stacking device 10 that separates and stacks the laminated wood lumber 9 that has been evaluated by inspection according to the evaluation. In Fig. 1, the laminated wood lumber 9 is separated into three types, plywood A 13, plywood B 14, and plywood C 15, but this is an example and is not limited to this.
The overall judgment controller 8 controls each inspection device and evaluates the plywood based on the inspection data, sorts it, and deposits it in a predetermined location.
These inspections, separations, and stackings are carried out continuously while transporting the laminated wood lumber 9 using the vertical transport conveyor 1, direction change conveyor 4, horizontal transport conveyor 5, and conveyor 11. The overall judgment controller 8 also controls these transport devices.

FIG. 2 is a front view showing the layout of the wood laminate inspection device.
A long-length end grain inspection device 30, a front and back inspection device 40, a short-length end grain inspection device 60, and a stacking device 10 are arranged.
The wood laminated lumber 9 is transported by a vertical conveying conveyor (the vertical conveying device of the long-length end grain inspection device 30 is not shown), a horizontal conveying conveyor 5A, and a horizontal conveying conveyor 5B, and passes through the long-length end grain inspection device 30, the front and back inspection device 40, and the short-length end grain inspection device 60, where it is evaluated and then handed over to the conveyor 11 of the stacking device 10, and depending on the evaluation, it is stacked on the stacking bar 12 at the stacked plywood A13, stacked plywood B14, or stacked plywood C15 locations.

3A, 3B, and 3C show a long end grain inspection device 30.
A long end grain imaging means 31 and a long end grain illuminator 32 are arranged on the side of the vertical transport conveyor 1 that transports the wood laminated lumber 9 in the longitudinal direction. On the left and right sides, a left long end grain imaging means 31L and a left long end grain illuminator 32L, and a right long end grain imaging means 31R and a right long end grain illuminator 32R are arranged, respectively.
A material detector 2 is provided on the conveying path of the vertical conveyor 1, and an encoder 3 is attached to the drive shaft. The material detector 2 detects the wood laminated lumber 9 being conveyed, and the conveying timing, the timing of irradiating the long end grain, and the timing of photographing obtained by the encoder 3 are stored and synchronized in time.

Figure 4 shows the inspection equipment for the front and back surfaces and the inspection equipment for short buttocks. (a) is a diagram of the inspection equipment for the long back surface, and (b) is a diagram of the inspection equipment for short buttocks.
As shown in FIGS. 1 and 2, the horizontal transport conveyor is divided into front and rear conveyors, and a front and rear inspection device 40 is installed between the front and rear conveyors, that is, horizontal carry-in conveyors 5A and 5B.
Short end grain inspection devices 60 are installed on both sides of the horizontal carry-in conveyor 5B.
A material detector 6 is installed above the horizontal carry-in conveyor 5A.

The front and back inspection device 40 is provided with a front surface imaging means 41 and a front surface illuminator 42 on the front surface side of the laminated wood lumber 9, and a back surface imaging means 44 and a back surface illuminator 45 on the back surface side of the laminated wood lumber 9.
The surface imaging means 41 includes a left surface imaging means 41L and a right surface imaging means 41R, which are arranged on the left and right. The surface illumination 42 includes a reflected light illumination 42A and an oblique light illumination 42B. Images obtained by these two types of illumination are captured by the surface imaging means 41.
The reflected light illuminator 42A illuminates the surface of the laminated wood lumber 9 so that reflected light is incident at a nearly right angle on the surface imaging means 41. The oblique light illuminator 42B illuminates the laminated wood lumber 9 at an angle (approximately 45 degrees in the illustrated example), and the surface imaging means 41 obtains an image using oblique light, which creates shadows due to unevenness.
The back surface imaging means 44 includes a left back surface imaging means 44L and a right back surface imaging means 44R. The back surface illumination 45 includes a reflected light illumination 45A and an oblique light illumination 45B. Images obtained by the back surface imaging means 44 are captured using these two types of illumination.
Both the reflected light illuminator 45A and the oblique light illuminator 45B are arranged to illuminate the front surface of the laminated wood material 9 as well as the rear surface imaging means 44.

The short wood end grain inspection device 60 is equipped with a short wood end grain imaging means 61 and a short wood end grain lighting device 62 .
A left short end grain imaging device 61L and a left short end grain illuminator 62L are provided on one side of the horizontal carry-in conveyor 5B, and a right short end grain imaging device 61R and a right short end grain illuminator 62R are provided on the other side. The short end grain imaging device 61 captures an image of the short end grain of the laminated wood lumber 9 based on the light illuminated by the short end grain illuminator 62.

An encoder 7 is attached to the drive shaft of the horizontal carry-in conveyor 5A. When the material detector 6 detects the laminated wood lumber 9 being conveyed, the conveyance timing, the timing of illuminating and photographing the front and back sides, and the timing of illuminating and photographing the ends of the grain on both sides obtained by the encoder 7 are stored and synchronized in time.
The horizontal carry-in conveyor 5A and the horizontal carry-out conveyor 5B are controlled to the same speed, so the photographing and storage of the short end grain inspection device 60 is also synchronized by the encoder 7 attached to the drive shaft of the horizontal carry-in conveyor 5A.

FIG. 5 shows an example of image generation obtained by imaging the front or back surface inspection.
In the front and back inspection device 40, the surface imaging means 41 (41L, etc.) captures an image based on the reflected light of two types of light, reflected light illumination 42A and oblique light illumination 42B, and generates an image 51. This image 51 is divided into a reflected light image separation 52A and an oblique light image separation 52B to generate separated images 52. The image 51 and separated images 52 are arranged in a row, with separated images obtained by line illumination or the like.
The separated image 52 is then processed sequentially to generate a reflected light image 53A and an oblique light image 53B. A magnified image 54 of the same knot is shown. The oblique light image 54B shows that the contour on the right side is less rounded than the reflected light image 54A, indicating that it is a hole. If this appears to have the same shape, it is flat and can be determined to be a live knot.

Figure 6 shows an example of an overlap image obtained by capturing images of long-length end grain inspection of laminated wood lumber.
This laminated wood lumber 9 is made up of nine layers. Part of the second layer of veneer overlaps the other. (a) is a schematic diagram, and (b) is an image taken with a long-length end grain inspection device. The surface is flat, but the veneers overlap inside, causing swelling in the overlapping areas and distortion in layers 2 to 4.
This wood laminate 9 is laminated so that the fiber directions are perpendicular to each other, so the fiber directions of the upper and lower layers are different, and the laminated state is clearly photographed.
By generating an image in conjunction with surface observation, traces of swelling or compression appear on the surface image, making it possible to determine the width and length of the overlap.

Figure 7 shows an example of a tunnel image obtained by capturing images of the end grain of a long piece of laminated wood lumber.
This laminated wood lumber 9 is made up of nine layers. Part of the veneer in the second layer is missing, leaving a hollow tunnel 21 between the layers. (a) is a schematic diagram, and (b) is an image captured using a long-length end grain inspection device. A slight depression appears on the surface, and the tunnel 21, the hole from the missing part, appears in the second layer as a black area where light is not reflected. In the image, it appears black because there is no light reflection.
The same applies to short-length end grain inspections.
By generating an image in conjunction with surface observation, a thin depression appears in the surface image, allowing the width and length of the tunnel cavity to be determined.

FIG. 8 shows a judgment flowchart of the inspection device for laminated wood lumber.
This judgment flowchart consists of a long cut end grain judgment flow 100, a front and back side judgment flow 200, a short cut end grain judgment flow 300, and an overall judgment flow 400. The grade of the laminated wood lumber is determined based on the long cut end grain judgment, the front and back side judgment, the short cut end grain judgment results, and a judgment that combines these.
The long and short cut grain determination flows are common to both the left and right sides, so only one is shown. The front and back side determination flows also have both the front and back sides, but since they are common, only the front side is shown.

1 shows a flow 100 for determining the end grain of a long piece of wood.
The left long end grain imaging means 31L, which images the long end grain on the left side, is a line sensor type that images the long end grain of the laminated wood lumber 9 at small width intervals while transporting it. The images obtained from the left long end grain imaging means 31L are combined by an image generation processing mechanism to construct video information of the entire long end grain and determine the long end grain.
This is done in the same way on the long end grain on the right.
In steps 104, 105, and 106, the CPU sets the judgment criteria, and while observing the captured image on a monitor, the CPU sets and accumulates the evaluation items and parameters, and the AI learns and establishes the evaluation criteria. After learning using the initial training data is completed, the accuracy of the AI judgment is improved as data is accumulated through implementation.
Information processing is performed by the left AI controller 33L and the right AI controller 33R.

101 Shading correction: Corrects unevenness in brightness and aberrations caused by the optical system in the line image obtained from the left long end grain imaging means 31L.
102 Memory storage: Line images are stored in memory in sequence.
103 Line image synthesis: Line images are synthesized in chronological order to generate an image of the entire length of the end grain.
104. Shake correction, warp/curl detection: Corrects shakes that occur when products (such as plywood) vibrate up and down while being transported on a belt conveyor, and corrects warping and curling that occurs in the product. Correction for warping and vertical vibrations of the belt conveyor will be explained later.
105 Tunnel detection: Detects tunnel-shaped defects such as insufficient original plate, insufficient core, and chips caused by partial loss of a layer.
106 Overlap detection: Detects overlapping layers.
107 Long end grain judgment: Judgment is made based on the number and size of tunnels and overlaps.
The long end grain determination signal is also sent to step 401.

The flow 200 for determining the front side is shown. The back side is also the same, so it is omitted here.
1 and 4, the surface has two imaging devices, a left surface imaging means 41L and a right surface imaging means 41R. If the wood laminated lumber to be inspected is short, it is possible to use only one imaging device.
The surface of the laminated wood lumber is imaged using left and right imaging devices with vertical reflected light illumination 42A and oblique oblique light illumination 42B. The illumination and imaging devices are synchronized by a synchronization circuit. The imaging device is a color line sensor type. Images are captured at minute intervals as the surface of the laminated wood lumber 9 is transported. The images obtained from the left surface imaging device 41L and the right surface imaging device 41R are combined by an image generation processing mechanism to generate an image 51 (see Figure 5), which is video information of the entire surface. The entire surface image is separated into two types of separated images 52 (see Figure 5): a reflected light image separation 52A and an oblique light image separation 52B. The separated images are sequentially processed to generate a reflected light image 53A and an oblique light image 53B, which are then evaluated for color and knot and hole presence to determine the surface.
A flowchart 200 for determining the front side is shown in Figure 8. This is also performed for the back side.
The judgment criteria for steps 211, 212, and 213 are set by the CPU, and the captured image is observed while being observed on a monitor, and evaluation items and parameters are set and accumulated, and the evaluation criteria are learned and established by the AI. After learning using the initial training data is completed, the AI will be able to make highly accurate judgments as data is accumulated through implementation.
Information processing is performed on the AI controller surface 43 and the AI controller back surface 46.

201 Imaging: The left side of the laminated wood material is photographed using the left surface imaging means 41L.
202 Shading correction: Corrects unevenness in brightness and aberrations caused by the optical system in the line image obtained from the left surface imaging means 41L.
203 Memory storage: Line images are stored in memory in sequence.
204-206: In the same manner as 201-203, the images taken by the right surface imaging means 41R are processed in parallel.
207 Generation of surface image: Calibration is performed by matching the position coordinates of the left and right images. An image 51 is generated, which is video information of the entire surface.
208 Image separation: Image separation by illumination. Separation into an image obtained from the light of the reflected light illumination 42A and an image obtained from the light of the oblique light illumination 42B.
209 Generate Reflected Light Image: Generate reflected light image separation 52A.
210 Oblique Light Image Generation: Generate oblique light image separation 52B.
211 Two-image comparison: The reflected light image separation 52A and the oblique light image separation 52B are compared to detect irregularities such as knots, holes, and cracks, as well as bark and resin spots. The number and area are also evaluated.
212 Oblique light image evaluation: Detects blue mold, discoloration, etc. Also evaluates the number and area.
213 Comprehensive judgment: The surface is judged based on the detections of 211 and 212.
The evaluation of 211, 212, and 213 is performed by first setting parameters in the control device (CPU) and then using teacher samples to make the AI learn. Furthermore, in the implementation stage, case studies are accumulated to improve accuracy.

FIG. 8 shows a flowchart 300 for determining the short end grain.
This flowchart 300 is the same as the flowchart 100 for determining the long end grain shown in Figure 8. This flow is performed on both the left and right short end grains, so one of them will be omitted.
The left short end grain imaging means 61L, which images the short end grain on the left side, is a line sensor type that images the short end grain at minute intervals while transporting the short end grain of the wood laminated lumber 9. The images obtained from the left short end grain imaging means 61L are combined by an image generation processing mechanism to construct video information of the entire short end grain and determine the short end grain.
In steps 304, 305, and 306, the control device (CPU) sets the judgment criteria, and while observing the captured image on a monitor, the evaluation items and parameters are set and accumulated, and the evaluation criteria are learned and established by the AI. After learning using the initial training data is completed, the accuracy of the AI judgment improves as data is accumulated through implementation.
Information processing is performed by the left AI controller 63L and the right AI controller 63R.

301 Shading correction: Corrects unevenness in brightness and aberrations caused by the optical system in the line image obtained from the left short end grain imaging means 61L.
302 Memory storage: Line images are stored sequentially in memory.
303 Line image synthesis: Line images are synthesized in chronological order to generate an image of the short end grain over the entire length.
304 Shake correction, warp/curl detection: Corrects shakes that occur when products (plywood, etc.) vibrate up and down when transported on a belt conveyor, and corrects warping and curling that occurs in the product. Correction for warping and vertical vibration of the belt conveyor will be explained later.
305 Tunnel detection: Detects tunnel-shaped defects such as insufficient original plate, insufficient core, and chips caused by partial loss of layers. The number and length are also evaluated.
306 Overlap detection: Detects overlapping layers. Evaluates the number and length.

FIG. 8 shows a flowchart 400 of the comprehensive judgment.
The overall judgment flow is a process in which the judgment signals obtained in the long edge grain judgment flow 100, the front and back judgment flow 200, and the short edge grain judgment flow 300 are linked together to judge the long edge grain, the short edge grain, and the front and back sides, and then these are combined to make an overall judgment.
For example, by sending a surface determination signal to the short end grain side, the surface side recess signal and the width and length of the tunnel 21 shown in Fig. 7(a) can be used to determine the width and length of the tunnel 21. Similarly, the surface side bulge signal and the overlap 20 shown in Fig. 6(a) can be used to determine the width and length.
The judgment of the end grain, which reflects and utilizes these surface signals and end grain signals, is done by setting the judgment criteria in the control device (CPU), observing the captured image while watching on the monitor, setting the evaluation items and parameters, accumulating them, and letting the AI learn and establish the evaluation criteria. After learning using the initial training data is completed, the accuracy of the AI judgment improves as data is accumulated through implementation.
This will also be achieved by setting initial standards and implementing AI learning to determine the surface, which reflects the signal from the end grain.
Furthermore, an overall judgment is made by combining the judgments of each element: long cut grain judgment, short cut grain judgment, and front and back surface judgment. This overall judgment also starts with manual learning to set the judgment criteria, and the accuracy of the AI judgment improves through accumulated learning during implementation.
Information processing is performed by the overall judgment controller 8 based on information from the controllers of the inspection devices for each part (AI controller left 33L, AI controller right 33R, AI controller front surface 43, AI controller back surface 46, AI controller left 63L, AI controller right 63R).

401 Long end grain determination and comparison of front and back coordinate axes: The signal from the long end grain determination (107) is compared with the signal from the surface determination (213) to align overlaps and tunnels on the target wood laminate using coordinate axes.
402 Long end grain, front and back determination: Based on the process of 401, the front and back images are taken together with the image of the long end grain to determine whether it is a long end grain.
403 Matching of short cut end grain data with front and back data: The signal from the short cut end grain determination (307) is compared with the signal from the surface determination (213) to align overlaps and tunnels on the target laminated wood material using coordinate axes.
404: Determining whether the short end grain is on the front or back side: Based on the process of 403, the images of the front and back sides and the long end grain are combined to determine whether the short end grain is on the front or back side.
405 Overall judgment: An overall judgment is made based on the judgment of the front and back surfaces in 213, the judgment of the long end grain in 402, and the judgment of the short end grain in 404. The plywood is graded as plywood A, plywood B, plywood C, etc., and piled up in the pile plywood A13, pile plywood B14, and pile plywood C15 locations shown in Figures 1 and 2.

The apparatus of the present invention will be described in more detail based on the schematic plan view of the laminated wood lumber inspection and sorting apparatus SS shown in FIG.
The inspection and separation device SS includes an inspection device A and a separation and deposition device B. The inspection device A and the separation and deposition device B are arranged in series and are controlled by a comprehensive judgment controller 8.
The laminated wood lumber inspection and sorting device SS continuously transports the laminated wood lumber 9, inspects the laminated wood lumber 9 during transport, evaluates and judges the laminated wood lumber 9 based on the inspection results, and stacks the laminated wood lumber 9 according to the evaluation.
The inspection devices A are arranged in succession in the order of a long wood end grain inspection device 30, a front and back inspection device 40, and a short wood end grain inspection device 60. The order of the inspection devices can be determined as appropriate.
The separating and piling device B is equipped with a piling device 10 that separates the wood laminated lumber 9 evaluated by inspection into three types of plywood A, B, and C based on the evaluation and piles them up.
These inspections, separations, and stacking are carried out continuously while transporting the laminated wood lumber 9 using the vertical transport conveyor 1, direction change conveyor 4, horizontal transport conveyor 5, and conveyor 11. Because the transport directions of long and short end grains change, a direction change conveyor is interposed between the vertical and horizontal transport.
The overall judgment controller 8 controls these transport devices and each inspection device, evaluation, sorting, and deposition.
The properties of the long end grain, the short end grain, and the front and back surfaces are photographed separately, but since it is necessary to evaluate the laminated wood lumber as a whole, the respective pieces of photographed information can be matched together based on the detection signals from the material detectors 2 and 6 installed on the conveying device and the information from the encoders 3 and 7 installed on the conveying device. A timer or other device can also be used to match the photographed information.

In the long end grain inspection device 30, while the laminated wood lumber 9 is being transported on the vertical transport conveyor 1, the detection signal detected by the material detector 2, the signal from the encoder 3, and the left and right imaging data signals (imaging data signals from the long end grain imaging means left 31L and the long end grain imaging means right 31R) are sent to the left and right AI controllers (AI controller left 33L and AI controller right 33R) to determine the position of the laminated wood lumber (the part of the laminated wood that the imaging data corresponds to is determined from the time elapsed since detection and the transport speed) and sent to the overall judgment controller 8. The long end grain judgment is performed using the imaging data of the long end grain according to the long end grain judgment flow 100.
This position information is also identified by the front and back inspection device 40 and the short end grain inspection device 60, and by comparing this position information, the imaging data of the long and short end grains and the imaging data of the front and back surfaces are integrated into a single piece of wood laminated material.

The conveying direction of the laminated wood lumber 9 is changed to an orthogonal direction by the direction changing conveyor 4, and the lumber is conveyed by the horizontal conveyor 5. The horizontal conveyor 5 is divided into a horizontal carry-in conveyor 5A and a horizontal carry-out conveyor 5B before and after the point where the surface is imaged by the surface imaging means. However, the conveying speeds of the horizontal carry-in conveyor 5A and the horizontal carry-out conveyor 5B are synchronized, and the position information of the laminated wood lumber 9 can be detected by the material detector 6 and the encoder 7.
A material detector 6 is installed above the horizontal carry-in conveyor 5A, which detects the conveyed laminated wood lumber 9. An encoder 7 is attached to the drive shaft of the horizontal carry-in conveyor 5A, and the location (coordinates) of the laminated wood lumber 9 can be identified from the signal from the material detector 6 and the conveyance signal from the encoder 7.

Between the horizontal carry-in conveyor 5A and the horizontal carry-out conveyor 5B, the front and back surface inspection device 40 is provided with a front surface imaging means 41, a reflected light illuminator 42A, an oblique light illuminator 42B, a back surface imaging means 44, a left back surface imaging means 44L, a right back surface imaging means 44R, a reflected light illuminator 45A, and an oblique light illuminator 45B. Images based on these two types of illumination for the front and back surfaces are obtained by the front and back surface imaging means. In this example, the imaging means are separated into left and right, with the front surface imaging means 41L, the front surface imaging means right 41R, the left back surface imaging means 44L, and the right back surface imaging means 44R arranged.
The image data obtained by the front and back imaging means undergoes front data processing in the AI controller front surface 43, and back data processing in the AI controller back surface 46, and the front side is determined according to the front and back side determination flow 200.

A short end grain inspection device 60 is arranged above the horizontal discharge conveyor 5B.
The short end grain of the laminated wood lumber 9 is illuminated by the short end grain lighting 62 and imaged by the short end grain imaging means 61. The lighting and imaging means are arranged on the left and right, and comprise a left short end grain imaging means 61L, a right short end grain imaging means 61R, a left short end grain lighting 62L, and a right short end grain lighting 62R.
The image data is processed by the left AI controller 63L and the right AI controller 63R. The short end grain is determined by taking into account data from the front and back sides. The short end grain is determined according to the short end grain determination flow 300.

Each of the long wood end grain inspection device 30, the front and back inspection device 40, and the short wood end grain inspection device 60 has an AI controller, but the overall judgment controller 8 is responsible for transmitting information between each device, overall control of the equipment, and overall judgment.
Using the judgments of the long length end grain judgment flow 100, the front and back judgment flow 200, and the short length end grain judgment flow 300, a comprehensive judgment of the laminated wood lumber 9 is made based on the comprehensive judgment flow 400, and the grade of the laminated wood lumber is determined.
Based on the grade information of the laminated wood lumber, the stacking device 10 is controlled to separate and stack the lumber according to grade.
The conveyors 11, 11, 11 of the stacking device 10 are controlled by the overall judgment controller 8, and the wood laminated lumber is stopped at each position, and by starting the stacking bar 12, it is stacked as stacked plywood A13, stacked plywood B14, and stacked plywood C15.

<About correction of inspection images>
Regarding surface images, [0025] describes that finely divided line images obtained by a line sensor are arranged as shown in Figure 5, and sequential processing is performed to generate a reflected light image 53A and an oblique light image 53B.
Similarly, for the image of the end grain, images divided into fine lines are arranged to generate a single continuous image of the laminated wood lumber for inspection.
When images of the end grain are taken, if the laminated wood lumber is warped or if the conveyor belt that transports it vibrates, the divided images can be bent and distorted, making it impossible to obtain a clean image for inspection. We propose a technology that corrects this and creates a single continuous end grain image, improving the accuracy of evaluation. The corrections described below can be applied to both long and short end grain images, so they will not be distinguished.
This image correction is included in 104 (correction of long grain ends) and 304 (correction of short grain ends) in the flow chart shown in FIG.

An example of a distorted image is shown in FIG.
For example, when the conveyor for transporting the laminated wood lumber vibrates, the laminated wood lumber also moves up and down, and the image of the butt section obtained by arranging the individual images (images taken by individual line sensors) 81 obtained by the line sensors at that timing becomes a wavy and distorted composite butt section image 70 as shown in Figure 9(a). Figure 9(b) shows a diagram of the distorted composite butt section image 70 divided into the individual line sensor images 81 that are the basis for the image.

FIG. 10 shows a method for obtaining a corrected image by performing correction based on the surface line or the center line.
The laminated wood lumber from which this distorted image was taken is a flat board of the same thickness, so the individual images obtained by the line sensor also have the same thickness. Therefore, by adjusting and aligning the individual images according to the front, back, or center, a straight, corrected end grain image 55 is obtained.

FIG. 10A shows an example in which the surfaces of the individual images are rearranged in accordance with the surface line 86 to produce a primary end grain corrected image 71, which is a corrected image.
Figure (1) shows the individual images x1, x2, ... xn obtained from the line sensor, arranged in chronological order. The image has a wavy, distorted middle portion due to the vibration of the belt conveyor. A surface line 86 is set to rearrange the image into a straight line, and a corrected image (primary edge-corrected image) 71 is shown in Figure (2) after the surfaces of the individual images x1, x2, ... xn are rearranged to match this surface line 86.
Since laminated wood is a flat board, if you rearrange a distorted image based on a certain point, you can obtain a straight image, which can be used as the corrected image for inspection. However, if there is a defect in the surface layer, the second layer will become the surface layer, and a depression will appear in the lower layer, which is incorrect. If a depression appears in the lower layer like this, a secondary correction will be performed.

The median values 85a of the individual images are rearranged to match the center line 85 to form the primary edge correction image 7.
An example of 1b is shown in FIG.
Figure (1) shows a diagram in which a center line 85 is set near the median 85a of each individual image, and Figure (2) shows an example in which the individual images are rearranged to align with the center line 85 to create the first-corrected image, the end grain corrected image 71b. However, if there is a defect in the surface layer, depressions will appear at the top and bottom, which is incorrect. If depressions appear at the top and bottom like this, a second correction will be performed.

An example of two-stage correction is shown in FIG.
When correction is made in accordance with the surface line, any defects on the front or back will appear in the corrected image as a recess 74 on the lower side. Figure 11(a) shows an example in which a recess has appeared in the lower layer of a primary end grain corrected image corrected based on the surface line. If there is a defect in the surface layer of end grain corrected image 71 corrected in accordance with the surface line 86, the defect will appear as a recess 74 on the lower side.
The cut edge corrected image 71 is a primary laminated wood lumber formed of five layers, from the surface layer (first layer) to the back layer (fifth layer). A recess 74 is generated in the lower layer for surface matching. This primary cut edge corrected image 71 is compressed in the length direction to create a reference image 87. The reference image 87 is made by superimposing the individual images, so it is possible to create a model of the cut edge of the laminated wood lumber.
FIG. 11( b ) shows a secondary correction in which the individual images are realigned to match the reference image 87 .
In the recess 74, the second, third, fourth and fifth layers are aligned, but the first layer is not aligned, so they are aligned below each other and the first layer (surface layer) is corrected as a missing portion, namely, a surface plate missing portion 72.

11A shows an example in which the primary end grain corrected image 71 corrected to match the surface line 86 has a defect in the surface layer. The defect appears as a recess 74 on the lower side.
The primary end grain corrected image 71 is a laminated wood lumber formed of five layers, from the surface layer (first layer) to the back layer (fifth layer). A recess 74 is generated in the lower layer for surface matching. This primary end grain corrected image 71 is compressed in the length direction to create a reference image 87. The reference image 87 is made by superimposing the individual images, so it is possible to create a model of the end grain of the laminated wood lumber.
FIG. 11B shows a secondary end grain corrected image 77 obtained by rearranging the images in accordance with the reference image 87 .
In the secondary end grain correction image 77, the second, third, fourth, and fifth layers match at the recess 74, but the first layer does not match, so they are lined up on the bottom and corrected as a missing top board part 72, where the first layer (surface layer) is a missing part.

FIG. 12 shows an example of secondary correction of a primary end grain corrected image 71b that has undergone center line correction.
12(a), if there is a defect in either the top or bottom layer, depressions 72 and 76b are created above and below the primary end grain corrected image 71b. Focusing on these depressions 72 and 76b and the areas before and after, and focusing on the top or bottom layer, corrections are made so that the same layers are continuous to generate a secondary end grain corrected image 77b.
In this case, the back layer of the recess 72 is the recess back board layer 76b, the left part is the front back board layer 76a, and the right part is the back back board layer 76c, which are all considered to be the same layer. Therefore, by adjusting the recess back board layer 76b downward, the missing part 72 of the surface board appears as a recess on the surface layer side, as shown in Figure 12(b). This is used for evaluation as the secondary end grain corrected image 77b.

Regarding laminated wood lumber, the layered veneers that form the layers of laminated wood lumber are shown schematically in FIG. 13(a).
This wood laminate is an example of thick plywood in which the grain directions of the veneers are laminated at right angles.
The wood fibers 27 are aligned in the direction of tree growth, and based on the relationship between the fiber direction and the cut surface of the tree, each surface of the layered veneer 22 shown in Figure 13(a) can be represented as a veneer flat grain 24, a veneer butt grain 25, and a veneer straight grain 26.
The common layer can be recognized by noting the veneer cut end 25 and the veneer straight grain 26 that appear in the cross section of the laminated wood material in which the veneers are layered.
The face veneer grain 26 has a smooth surface with parallel fibers, so light is reflected in the image in many cases. The veneer end grain 25, which runs across the fibers, has countless small irregularities that cause light to be diffused, so the difference is clearly visible on the image screen.
An example of a five-layer laminated wood material 9a, in which five veneers are laminated so that the wood fiber directions are perpendicular to each other, is shown in Figure 13 (
The odd-ply veneer 22, which comprises the odd-numbered layers 1, 3 and 5 from the front surface 91a side, has fibers running in the longitudinal direction, with the veneer end grain appearing on the short end grain 93 side and the veneer straight grain appearing on the long end grain 92 side.
In the even-numbered layer veneer 23 constituting the even-numbered second and fourth layers, the straight grain of the veneer appears on the short end grain 93 side and the end grain of the veneer appears on the long end grain 92 side.
In the illustrated five-ply laminated wood material 9a, a missing portion 22 is shown in the first layer. In the second correction, it is easy to compare with the reference image and adjust to a common layer structure.
In addition, in LVL, which is made by stacking boards such as laminas, the quality of the laminas is managed individually, so it is rare for parts to be missing.

In this way, if there is distortion in the primary image synthesized from the images of the end grain obtained from the line sensor, a corrected image (primary corrected image) is generated by arranging the individual images in a continuous straight line, focusing on the top layer, back layer, center, etc., and this is used as the image for inspection. If there is a depression in this corrected image, secondary correction is performed, such as arranging individual images in the depression and the reference image, or comparing the front and back layers and adjusting them by focusing on the common layer, to generate a secondary end grain corrected image and use it as the image for inspection.
By performing the first and second corrections, accurate evaluation can be performed, and the evaluation precision is improved.
This correction technique can be applied to both long and short end grains.

1 Vertical conveying conveyor 2 Material detector 3 Encoder 4 Direction change conveyor 5 Horizontal conveying conveyor 5A Horizontal carry-in conveyor 5B Horizontal carry-out conveyor 6 Material detector 7 Encoder 8 Overall judgment controller 9 Wood laminated material 91 Surface 91a Front surface 91b Back surface 92 Long end 93 Short end

10 Stacking device 11 Conveyor 12 Stacking bar 13 Stacked plywood A
14. Stacked plywood B
15. Stacked plywood C
20 overlap 21 tunnel 22 layer veneer 24 veneer grain 25 veneer butt end 26 veneer straight grain 27 wood fiber

30 Long wood end grain inspection device 31 Long wood end grain imaging means 31L Long wood end grain imaging means left 31R Long wood end grain imaging means right 32 Long wood end grain lighting 32L Long wood end grain lighting left 32R Long wood end grain lighting right 33L AI controller left 33R AI controller right

40 Front and back inspection device 41 Front surface imaging means 41L Front surface imaging means left 41R Front surface imaging means right 42 Front surface illumination 42A Reflected light illumination 42B Oblique light illumination 43 AI controller front surface 44 Back surface imaging means 44L Back surface imaging means left 44R Back surface imaging means right 45 Back surface illumination 45A Reflected light illumination 45B Oblique light illumination 46 AI controller back surface

51 Imaging 52 Separated image 52A Reflected light image separation 52B Oblique light image separation 53 Image generation 53A Reflected light image generation 53B Oblique light image generation 54 Image generation Enlarged photograph 54A Enlarged reflected light image photograph 54B Enlarged oblique light image photograph 60 Short end grain inspection device 61 Short end grain imaging means 61L Short end grain imaging means left 61R Short end grain imaging means right 62 Short end grain lighting 62L Short end grain lighting left 62R Short end grain lighting right 63L AI controller left 63R AI controller right

70 Composite end grain images 71, 71b Primary end grain corrected image 72 Missing portion of top board 74 Recess 75 Top board layer 76 Back board layer 76b Recess back layer 76a Front back layer 76c Back back layer 77, 77b Secondary end grain corrected image 81 Individual image 85 Center line 85a Median 86 Surface line

Claims (9)

This is an edge inspection device that uses the reflected light of edge irradiation light irradiated on the edge of wood laminated lumber,
The end grain inspection device is
a long edge grain inspection device that is equipped with a long edge grain illumination and a long edge grain imaging means, and that determines the properties of the long edge grain from the long edge grain image, which is an image for long edge grain inspection obtained by rearranging and correcting the images taken by the long edge grain imaging means in the thickness direction of the wood laminated lumber;
The short-length end grain inspection device is provided with a short-length end grain illumination and a short-length end grain imaging means, and the image taken by the short-length end grain imaging means is rearranged in the thickness direction of the wood laminate and corrected to obtain a short-length end grain corrected image, which is an image for short-length end grain inspection, and determines the properties of the short-length end grain from the short-length end grain image.
1. An inspection device for laminated wood lumber , comprising : An inspection device for wood laminated lumber as described in claim 1, characterized in that if there is a part in the long-length end grain corrected image and/or the short-length end grain corrected image where the thickness differs before and after the correction , a second long-length end grain corrected image or a second short-length end grain corrected image which has been corrected to be common to the layer of that part by comparing it with the reference long-length end grain image or short-length end grain image is used as the image for inspection. An inspection device for laminated wood lumber, comprising: a surface inspection device that utilizes the reflected light of light irradiated onto the surface of laminated wood lumber; and an end grain inspection device that utilizes the reflected light of light irradiated onto the end grain of the laminated wood lumber described in claim 1 or 2 . 4. The inspection device for laminated wood lumber according to claim 3 , wherein the surface inspection device is provided with inspection devices for both the front and back surfaces. 4. The inspection device for laminated wood lumber according to claim 3, wherein a comprehensive judgment is made based on the surface judgment made by the surface inspection device and the end grain judgment made by the end grain inspection device. 3. The inspection device for laminated wood lumber according to claim 1, wherein the laminated wood lumber to be inspected is any one of plywood, LVL, LVB, CLT, and thick plywood. 6. The inspection device for laminated wood lumber according to claim 5 , wherein the laminated wood lumber to be inspected is any one of plywood, LVL, LVB, CLT, and thick plywood. 7. An inspection and sorting device for laminated wood lumber, comprising: the inspection device for laminated wood lumber according to claim 6 ; and a stacking device for sorting and stacking the lumber based on the inspection results of the inspection device. 8. An inspection and sorting device for laminated wood lumber, comprising: the inspection device for laminated wood lumber according to claim 7 ; and a stacking device for sorting and stacking the lumber based on the inspection results of the inspection device.