CA2933490C - Charger scanner system - Google Patents

Abstract

A measurement system and method for computing a geometric center of a log block in three dimensions is provided. The system comprises a charger apparatus and at least one block scanner mechanism. The scanner block mechanism includes an illumination source for projecting light along the length of the surface of the log block, forming a plurality of scan points and an imaging device for receiving the plurality of scan points that are reflected from the surface of the log block. The system further comprises processing unit that compiles information from the imaging device representative of a three-dimensional image of the log block. The processing unit calculates an optimum yield axis from the three-dimensional image.

Description

CHARGER SCANNER SYSTEM
Field of the Invention The present invention concerns apparatus and methods for analyzing the geometry and other attributes of a log block in three dimensions, more specifically generating a plurality of scan points for collection within a measurement system used to compute the optimum yield axis of the log block.
Background of the Invention Veneer sheets are generated from tree logs. Logs are ultimately positioned in a veneer lathe charger and it is desirable to calculate, as close as possible, an optimum yield axis of the log veneer when, subject to peeling. Current industry practice is to acquire measurements surrounding a log circumference at multiple stationary locations spaced along the length of the log. The measurements are used to obtain only an approximate center of the log in both a lathe charger and veneer lathe prior to the peeling operation.
The logs, blocks, or log blocks as they are often designated can exceed 40 inches in diameter and over 106 inches in length when received by the veneer lathe. When in the lathe, the block is turned at a high speed and is engaged along its entire length by a lathe knife, which strips the veneer away, ideally in uninterrupted sheets. The sheets are typically 0.015"-0.25" inches thick, and the log is typically turned until the diameter of the block is less than six inches. Breaks in the sheets of the veneer occur most often when the spindles turning the block are not located about an optimum yield axis. Such breaks equate to waste resulting in undesirable cost and expense. Significant research and development has been allocated to improving equipment and methods for extracting veneer from log blocks. The assignee of the present invention has several patents involving such improvements.
Such patents include, United States Patent Nos. 6,116,306, issued September 12, 2000 and 4,884,605, issued December 5, 1989, which describe noncontact scanning using triangulation measurement techniques and multiple single point laser scanners for determining diameter cross-sections of the log block.
Quantities of 5 to 32 independent single point scanners are current common practice for a typical eight foot lathe charger. Installation, alignment, cleaning, and maintenance of large quantity independent scanners are costly and inefficient. Although the techniques described in the aforementioned patents for noncontact scanning an approximate center axis is generally effective, the art lacks the ability to calculate an optimum yield axis. Thus, less than a maximum yield is achieved by employing the techniques currently practiced in the industry as described in the above patents. Anything less than an accurate calculation of the log's optimum yield axis creates waste resulting in costs to the overall product.
Summary of the Invention The present invention is a new and improved method and apparatus for scanning and analyzing a log block profile for determining an optimum yield axis using a series of block profile scanners.
According to one example embodiment, four block profile scanners are employed.
Each block profile scanner is capable of measuring 512 scanned distances.

2 In summary then, the present invention may be characterized in one aspect as including a measurement system for computing a geometric center of a log block in three dimensions where the system includes a charger apparatus; at least one block scanner mechanism having an illumination source for projecting light along the length of the surface of the log block forming a plurality of scan points and an imaging device such as a two-dimensional pixel array for receiving the plurality of scan points that are reflected from the surface of the log block; and a processing unit such as a computer that compiles information from the imaging device representative of a three-dimensional image of the log block, the processing unit calculating an optimum yield axis from the three-dimensional image. The system may include a plurality of block scanners mechanisms in a spacial alignment such that some of the plurality of scan points from adjacent scanners overlap.
In one embodiment the imaging device includes a two-dimensional pixel array having a scan density of at least 512 x 512 pixels and is a charged couple device, that is includes two dimensional CCD arrays.
The imaging device may be a complementary metal oxide semiconductor device.
The block scanner may be a triangulation type measuring scanner. The processing unit may further advantageously include a scan overlap algorithm for analyzing the overlapping scan points of adjacent block scanners, for example wherein the scan overlap algorithm uses the scan point having the longer distance of any two or more overlapping scan points when compiling the information representative of the three-dimensional image of the log block to calculate the optimum yield axis. Yet further advantageously, the processing unit may further include a range threshold algorithm such that any scan point having a value less than the threshold is negated when compiling the information representative of the three-dimensional image of the log block to calculate the optimum yield axis.

3 =
The method for scanning a profile of a log block according to a further aspect ofthe present invention may include the steps of:
a) transmitting illumination from an illumination source located within at least one block scanner along the length of the surface of the log block;
b) receiving reflected illumination from along the length of the surface of the log block in the block scanner onto an imaging device, the reflected illumination forming a plurality of scan points;
c) rotating the log block about its axis while transmitting illumination and receiving the reflected illumination;
d) generating a plurality of scan paths from the scan points collected by the imaging device;
e) triangulating the scan points for each of the scan paths to produce a numerical model of the block; and f) computing the geometric center of the block from the numerical model of the block.
The method may further include a preliminary and secondary scanning process, and the log block may remain stationary during the preliminoty scanning process.
The method may include filtering out the scan points having a value less than a range threshold value when triangulating the scan points for each ofthe scan paths to produce a numerical model of the block, or filtering out overlapping scan points captured by more than one block scanner when triangulating the scan points for each of the scan paths to produce a numerical model of the block.
Brief Deacription of the Drawings The foregoing and other features of the present invention will become apparent to one skilled in the art to which the present invention relates upon consideration of the following description of the invention with reference to the accompanying drawings, in which:

4 Figure lA is an isometric view of a scanner used in the prior art;
Figure 1 is a front elevation view of a charger and scanners where the log block is in a first or rest position for a preliminary scanning process;
Figure 2 is a side elevation view of Figure 1 and of a veneer lathe;
Figure 3A is an isometric view showing the operation of a block profile scanner in one example embodiment;
Figure 38 is a diagrammatical view of one example embodiment of a block profile scanner;
Figure 4 is a front elevation view of the charger and scanners where the log block is in a second or yield scan position;
Figure 5 is a partial perspective view of Figure 4;
Figure 6 is a linear representation of a block circumference through 3600 as shown in Figure 5;
Figure 7 is a partial front elevation view of a charger positioning a block such that the scanners are used for calculating an optimum yield axis;
Figure 8 is a magnified view of a block exposed to overlapping scanner regions encountering a shadowing condition in the block profile;

5 Figure 9 is a flow diagram representing a logic process of analyzing a shadowing condition in a log block profile; and Figure 10 is an illustration representing a filtering process used when scanning a block profile.
Figure 11 is a diagrammatic view of a block in cross section and the corresponding representations of the r, bõ, and h dimensions.
Detailed Description of Embodiments of the Invention Turning to the drawings, a lathe charger apparatus is shown in phantom generally at 10 in Figures 1 and 2. The lathe charger 10 holds a log block 12 in a rest or first position supported by a first vee 13 and a second vee 14. The block 12 comprises a first end 15 and second end 16, as supported by the vees upon loading into the charger apparatus 10 from a conveyor (not shown). Depicted in Figure 1 are block profile scanners or cameras 20a - 20d suspended above the log block 12 by a fixture 11.
Although four scanners 20a - 20d are shown, it should be appreciated by those skilled in the art that any number of scanners can be used. A block profile scanner 20 is a triangulation type-measuring scanner =
capable of capturing a two-dimensional image. The block profile scanner can be for example, a type of scanner that employs a charged couple device (CCD) or complementary metal oxide semiconductor (CMOS) based imaging technology. A suitable block profile scanner will have an illumination source that is projected along the length of the surface of the log block 12, reflecting a surface image profile of the block surface along its length onto a high density two-dimensional pixel array 26.
Figure 1 A an isometric view of a scanner system used in the prior art. The scanner system includes an illumination source projecting light along the circumference of a block. The

6 camera or scanner in Figure 1 A is limited to imaging the surface of the block at a single circumferential location along the block.
Figure 3A illustrates one example embodiment of a scanner system having an illumination source 23 separated from a camera 32. The illumination source 23 projects a light path 24 upon the block 12. The light path 24 is reflected from the top surface of the block forming a reflected image 29 that is received by a lens 25 on the camera 32. An analysis of the reflected image 29 is performed by a processor (not shown) that can either be internal or external to the camera 32 or scanner 20. A separate example embodiment is shown in Figure 3B. Structures of Figure 3B that are the same or similar to structures of Figure 3A and are labeled with the same reference number with the addition of a prime.
Figure 3B features an illumination source 23' and receiving lens 25' contained within a scanner 20'. The reflected image 29' is ultimately projected through the lens 25' onto the high density two-dimensional pixel array 26. In one embodiment, the pixel array has a scan density resolution of 512 x 512 pixels per scanner.
An example of a suitable commercial block profile scanner includes a camera made by IVP
Integrated Vision Products Inc. under model number NP Ranger M20 OEM-1.
Further discussion on this type of scanner technology is disclosed in U.S. Patent No. 6,313,876 issued to Eklund.
The scanner 20' of Figure 3B further includes a housing 21 with a window 22 from which the illumination source 23' projects a light path 24' upon the surface of the block 12. The illumination source 23' can be one or more light emitting diodes or laser diodes lensed to project a narrow line of light. The light path 24' results in a reflected image 29' of the block surface and is received through a lens 25' located in the scanner 20'. The lens concentrates the reflected image 29' upon the two-dimensional pixel array 26. The data image captured on the pixel array is then

7 analyzed and converted into a value that corresponds to a linear scan distance from the scanner 20 that is associated with a specified scan point or position r. along the block 12.
Employing a 512 x 512 pixel array, approximately 512 scan distances are measured by each scanner along the profile of the block 12 at rotational increments ofthe log block at 1.50 or less. Each scan distance is correlated to the simultaneous angle of log block rotation and to the simultaneous distance from scanner face to axis of the log block rotation, thereby providing the three-dimensional data for calculating an optimum yield axis.
Preliminary Scanning Operation A preliminary scanning process occurs while the block 12 is in a first position in order to calculate an approximate center 30, as shown in Figure 2. During the preliminary scanning operation, the four block profile scanners 20a 20d analyze and scan the block 12 from the first end 15 to the second end 16, as best seen in Figure 1.
Each of the block profile scanners generate 512 scan points represented by the values n, = 1'0 in Equation (1) and shown in Figure 3B. The values rõ
correspond to a particular pixel in the pixel array 26 that is analyzed and converted into a corresponding scan distance. Each scan distance corresponding to each scan point is collected and stored in a central processing unit (CPU) or computer 40. A variable h of Equation (1) represents a known distance from the scanners less the distance from the calculated scan point. Knowing the values for hand each scan distance corresponding to each scan point r0 along the profile 12 of the block allows for a distance b. to be calculated, which represents the approximate center 30, as demonstrated by the equations and Figure 11.

8 h=r+b Equation (1) rappwõ b. cos (45 ) Equation (2) b = h /1 + cos (45 ) Equation (3) $econdary Scanntng Operation for Calculating the Optimum Yield Axis After the completion of the preliminary scan and the calculation of the approximate center 30, a pair of charger spindles shown in Figure 4 as 17a and 17b, are independently moved into engagement with the first and second ends of the log block 12 at the approximate center 30 calculated during the preliminary scanning operation. A pair of hydraulic cylinders 18a and 18b lift block 12 from the first rest position to a second or yield position, as best seen in Figures 4, 5, and 7.
While block lifts from first rest to yield position or while in the yield position, log block 12 assumes an orientation for determining a precise three-dimensional geometric configuration from which the optimum yield axis can be calculated by rotating the log block 12 via charge spindles 17a and 17b. Simultaneously, four light paths represented by 24a - 244 are generated by the illumination source 23 in each respective scanner and are projected along the length profile of the block 12. Each light path contains 512 scan points represented by rõ that provide a reflected image on the pixel array 26, which is equated to a corresponding scan distance to the block 12. The four light paths 24a ¨ 24d combine to form a single scan path 31. As the block rotates, multiple scan paths 31a, 31 b... are generated along a longitudinal direction represented by X in Figures 5 and 6 through the 360 degree rotation. Log block 12 is typically accelerated into a spin col at an approximate rate of 2 revolutions per second and using the described block profile scanners allow for scan paths to occur every 1.5 degrees over the 360 degree rotation of the block 12, thereby generating at least 240 scan paths.
The number and frequency of the scan paths could be even greater during the acceleration and deceleration of the block rotation, most typically as the rotational speed decreases. By using four block profile

9 scanners each having 512 x 512 two-dimensional pixel array and generating a scan path along the block approximately every 1.5 degrees results in approximately 419,520 scan points rõ that are measured by the scanner and thus provide a corresponding scan distance SD for each scan point, as shown in Figure 7. Such a significant number of scan points along multiple scan paths provides an elevated scan density, allowing a clear three dimensional analytical depiction of a log block, including detection of any excrescences or discontinuities 28 in the block that often result from knots in the wood. Typically, the larger to the mass of the block 12 the slower the rotational speed, resulting in an even greater number of scan point rõ than the 419,520 points discussed above.
Knowing the scan distances SDõ based on pixel data from the scan points r,, provide corresponding scan point radii "Re" that collectively form the optimum yield axis of the block 12 represented by line 50 in Figure 7. For example, knowing SD and the height H
from the center of the spindle 17 to the corresponding scanner 20, and the angle that a given light beam 24 is projected from scanner center represented by allows the scan point radius 11õ
to be calculated, which corresponds to a radius of the block at that particular scan point.
Although H is a variable distance it is known value, since the distance traveled for each spindle 17 can be tracked by the CPU 40 by, for example encoders (not shown). The scan point radius Rõ is the distance from the surface of the block 12 to the optimum yield axis 50 for each scan point rõ.
This is shown and calculated for scan point radius 11512 that corresponds to scan point 1.512 in Figure 7 and Equation (4).
H512 = H - X
X = cos (13) * SD
H512 = H - cos() * SD Equation (4) The locus of the scan point radii Rõ from the data stream generated by 512 scan points from each scanner along every 1.5 degrees of rotation of the block 12 provides a precise three dimensional shape of log block 12. The three dimensional shape provides approximately 419,520 scan point radii from which the optimum yield axis 50 is calculated. Once the optimum yield axis 50 is computed, such location is used to position block 12 for the highest yield value in the removal of veneer by receiving a pair of pendulum clamps for transferring the log block 12 into a veneer lathe 60 shown generally in phantom in Figure 2. The process executed by the veneer lathe 60 is described in further detail in the '306 patent.
Scan Path Overlap 1.0 Spacing of the scan points I.. from each scanner along log block length varies with distance from scanner to log surface, typically ranging from 0.050" to 0.100"
over the operating range of each scanner. When scanners overlap the result is more dense or closer spaced data and sometimes (with reference to Figures 7 and 8) two X readings may coincidentally apply to the same Y position. An optimizing algorithm 100 will use the smaller indicated radius for calculating optimum yield axis 50. An added advantage of overlap is more accurate surface definition by eliminating shadowing that may result from surface protrusions and recesses.
Each scan distance SD is converted into distance X (perpendicular distance from face of scanner to stuface of log as shown in Figures 7 and 8) and distance Y
along length of log 12 to a horizontal center location. For each scanner, Y is first expressed as distance from vertical centerline of the scanner, negative to right of center and positive to left.
The scanner interface computer 40 combines data from all four scanners and passes all range readings to the optimizing computer in the form of X (perpendicular distance from face of scanners to surface of log) and Y
(distance along length of log) with Yr.:iat center of material flow, negative Y to right and positive Y to left. For example, (assume X is constant 63"), the last SD at each end of a block 100" long would be reported to the optimizing computer as X = 63", Y = -50" right end and X = 63" and Y
+50" left end.

Shown in Figure 1 are three overlap regions represented by 27a, 27b, and 27c.
Each overlap region 27 results from each respective scanner light path 24 extending beyond the adjacent scanner's light path. The amount of surface area of the block 12 subjected to the overlap regions 27 vary depending on the diametrical size and location of the block.
The overlapping regions are smaller as the diameter of the block 12 increases or the closer the proximity of the block to the scanner, as depicted when comparing Figure 1 (the block located in a preliminary scan position) with Figure 4 (the block located in a secondary scan position).
The overlap regions 27 provide an advantage in analyzing areas susceptible to a shadowing condition along the block resulting from projections or protuberances along the block profile as best seen in Figure 8, which is a magnified portion of a log block 12. The overlapping regions, for example in Figure 8 using scanner 20b and 20c provide two different scan distance values for scan point I'. represented by SD1 and SD2, respectively. The optimization algorithm 100 constructed in programs or source code resolves which scan distance to use for overlapping scan points, which is represented generally, by the flow chart in Figure 9.
Referring now to Figure 9, the optimization algorithm source code 100 is initiated at 110. The scan distance reflected onto the pixel array 26 for each scanner is assigned a value at 120 and assigned a coordinate position at 130 relating to the value 120. A
decision point 140 evaluates whether the coordinate points are within an overlapping region and assigned the same position. If the points do not overlap the scan point radius R, of the block 12 is calculated for that particular scan point I-. at 145. If the scan points do overlap and are assigned the same position, a comparison is made at 150. The scan distances are compared for each common scan point at 150.
The scan point having the greater scan distance is selected for calculating the block scan point radius R. at 160. Referring again to Figure 8, SDI is greater than SD2, therefore the value for SDI will be used to calculate the scan point radius R. for scan point rA in accord with the program 100. As a result, the programming and overlapping regions 27 provide a smaller diameter and more accurate image of the log block 12 when encountering shadowing conditions.

Steam Penetration Filtering The scanning environment frequently includes airborne steam, or any other airborne material, of random intensity between the camera or scanner 20 and the block 12 being scanned. The steam is typically emitted from the block 12 being scanned and/or from peeling veneer from the previously scanned block. Even though laser illumination and reflection from the block surface penetrate the steam, random reflections from the steam can cause multiple extraneous range readings.
Figure 10 is an illustration representing a filtering process used when scanning a block profile.
The cameras 20 are scanned from the farthest range to closest. Only the farthest range reading represents the block surface thus only the farthest reading is used for calculating optimum yield axis. A
threshold 70 is provided to remove background ambient light and allow for accurate range measurement. Stated another way, if the scan distance measures an object such as random reflections, airborne material or steam clouds 72, the measured scan distance will be smaller than the threshold 70, and as such, the scan distance measurement is dismissed or filtered from the central processing unit or computer 40 when calculating the optimum yield axis 50. The scope of the claims should not be limited by the embodiments set forth in the examples, but should be given the broadest interpretation consistent with the description as a whole.

Claims (7)

WHAT IS CLAIMED IS:

1. A method for scanning a profile of a log block comprising:
transmitting illumination along the length of the surface of said log block;
receiving reflected illumination from a plurality of locations along the length of the surface of the log block in a plurality of imaging devices, the reflected illumination forming a plurality of scan points that correspond to said locations, wherein the imaging devices include a first and a second imaging device with overlapping fields of view;
rotating said log block about an approximate center axis of the log block while transmitting the illumination and receiving the reflected illumination;
generating a plurality of scan paths from the scan points collected by the imaging device;
triangulating and filtering the scan points for each of the scan paths to produce a numerical model of said log block, wherein filtering the scan points includes selecting, from among a pair of scan points for a given location within the overlapping fields of view, only one of the scan points to represent said given location; and computing the geometric center of said block from the numerical model of said log block.

2. The method of Claim 1, further comprising performing a preliminary scanning process before rotating said log block, wherein performing the preliminary scanning process includes scanning the log block with the imaging devices to determine the approximate center axis of the log block based on said scanning.

3. The method of Claim 2, further including moving a pair of charger spindles into engagement with opposite ends of the log block at the approximate center axis after the preliminary scanning process and moving the log block toward the imaging devices before rotating the log.

4. The method of Claim 1, wherein triangulating the scan points includes determining a scan distance for each of the scan points, and wherein filtering the scan points further includes removing or disregarding scan points with a scan distance that is less than a range threshold value.

5. The method of scanning a profile of a log block in Claim 1 wherein triangulating the scan points includes determining a scan distance for each of the scan points, and wherein filtering the scan points captured by more than one block scanner when triangulating the scan points for each of the scan paths to produce a numerical model of said block.

6. The method of scanning a profile of a log block in Claim 5, wherein filtering the scan points further includes removing or disregarding scan points with a scan distance that is less than a range threshold value.

7. The method of any one of Claims 1-6, wherein transmitting the illumination along the length of the surface of said log block includes using a plurality of light sources to transmit corresponding light paths that overlap at adjacent ends along the length of the surface of said log block.