Multi-view x-ray imaging of logs

Details Multi-view x-ray imaging of logs Multi-view x-ray imaging of logs

2002-09-20

2004-06-29

Bruce, David V (Department: 2882)

X-ray or gamma ray systems or devices

Specific application

Absorption

C378S004000, C378S010000, C378S062000, C378S901000

Reexamination Certificate

active

06757354

ABSTRACT:

CROSS-REFERENCE TO SOFTWARE APPENDIX
Per the requirements of 37 C.F.R. §1.96, a Computer Program Listing Appendix is a part of this disclosure and is incorporated by reference. The Computer Program Listing, Appendix A, is contained on a total of two CD-ROM's, consisting of a first copy, “Copy 1,” and a duplicate copy, “Copy 2.” Appendix B, which is also part of the present specification, contains a list of the files contained on the compact disk(s), and is attached herewith. The files of Appendix A form source code of computer programs and related data of an illustrative embodiment of the present invention. The CD-ROM is in IBM-PC format and compatible with MS-Windows.
A portion of the disclosure of this patent document contains material, which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the U.S. Patent and Trademark Office patent files or records, but otherwise reserves all copyright rights whatsoever.
BACKGROUND
1. Field of the Invention
This invention relates to x-ray imaging of internal structures of solid bodies, and the particular examples relate to the determination of the location and geometry of defects internal to logs.
2. Discussion of Related Art
The variability of logs in terms of shape and quality is a major challenge for modern sawmills. To optimize the output of each individual saw log, information as to the shape and quality has to be known prior to sawing. Two different strategies are used to optimize the log breakdown. In European softwood mills the logs are pre-sorted according to dimensions and sometimes quality. Usually, optical scanners provide the data on which the sorting is based. Then, each sorted group of logs is sawed with the identical sawing pattern. In most cases the sorting is done while the bark is still on. A limitation of optical scanners is that they only measure the external shape of the log, whereas the more relevant shape under bark remains unknown. The dimensions under bark are often estimated using a statistical bark thickness function. Variations in bark thickness thus cause sorting errors. The other strategy is to optimize the sawing pattern of each individual log. This is common in mills processing higher-value logs such as hardwood or large-diameter softwood logs. In both cases, information about the external shape as well as the internal quality—the locations and distributions of defects—is needed to make the best decision.
Internal log scanners are a class of scanners, which provide information about the interior of the logs. X-ray imaging has for several years demonstrated the ability to reveal the location and geometry of internal structures in logs, which will serve as a means for improving the yield of quality lumber after sawing of the logs. Knowledge of the internal structures, such as knots, will improve the sawing decision by rotating the log into a known orientation upon presentation to the primary breakdown in the sawmill. In some cases like in pruned planted pines (Radiata Pine, Southern Yellow Pine), the knot core of the log will be measured, and thus the yield of high-quality lumber free of knots from a given log is improved.
It remains a challenge to the timber industry to develop and implement algorithms for most effectively analyzing the data generated by x-ray sources and detectors. One approach is set forth in Aune et al. in U.S. Pat. No. 5,023,805, where a three-view x-ray apparatus is disclosed. This system detects the knots in the x-ray views and applies a convex hull method to determine the knot geometry. The application of the system is optimization of the primary breakdown on a band saw. In Pietikainen et al., U.S. Pat. No. 6,157,698, an evidence value representative of the presence of a knot in a particular volumetric element of a log is calculated, and the evidence values of mutually associated elements are combined to produce an aggregate evidence value. The evidence value is used to sort logs according to their apparent quality.
The methods disclosed in the prior art suffer from a number of disadvantages. For instance, the methods cannot distinguish the outline of the bark from the shape of the log under the bark, since a simple laser profile scanner is used to detect the shadow edges of the log. Therefore, the true shape of the log under bark cannot contribute to the selection of a sawing decision or to a measurement of the density of the wood. Also, the convex hull method described in U.S. Pat. No. 5,023,805 does not effectively address the situation in which several defects are located in the same log cross section. Therefore, the ability of the prior art approaches to identify and locate defect information corresponding to knots in the wood is limited. Hence, an improved approach is desirable.
SUMMARY OF THE INVENTION
The present invention offers significant improvement in terms of modeling the shape of the heartwood/sapwood boundary and the bark thickness. This modeling is useful in terms of describing the geometry of the log, but has a further advantage in that it allows for more precise mapping of internal defects in the log, such as knots. A further improvement lies in the incorporation of knowledge of the morphology of the tree anatomy into the mapping method, to discriminate between knots and clear wood.
Embodiments of the invention make use of two or more x-ray sources, which emit radiation, which impinges upon a solid body whose internal structures are to be characterized. Each of the x-ray sources has a corresponding sensor, which is an array of individual detectors arranged side-by-side and on the opposite side of the body from the source, and in a circumferential pattern so that the individual detectors form an arc whose center is at the location of the source. In the example, the solid body is a wood log, and the internal structures of interest include the heartwood, sapwood, knots and other defects, as well as inclusions such as metal spikes or nails. The radiation is transmitted through the body as a function of its density, such that higher density structures, such as knots, attenuate the radiation more substantially than lower density areas, such as heartwood.
By measuring differences in the transmitted intensity between adjacent detectors, the ray paths which just intersect a boundary between two regions of different density can be determined, and a tangent line is calculated which just intersects a boundary between the regions. The differences in the higher order features of the transmitted intensity, such as texture and gradient, may also contain information as to boundaries between defects and clear wood, or between different categories of clear wood. Knowing the location of the source allows an equation for the tangent line to be calculated, and an estimate of an ellipsoidal shape is determined, which would approximate the detected signal. The method assumes an arbitrarily oriented ellipse as the model shape, and the tangent lines are fitted to the model. An improved model shape is obtained by using an interpolation correction between the measured attenuation values and the model predictions.
By transporting the log past the sources and sensors in a direction perpendicular to the paths of the x-ray radiation, a longitudinal projection view of the log is generated, with each detector imaging a single chord through the body of the log. Low frequency (slowly varying) changes in this detected signal are indicative of variations in the width of the log, or its curvature. Higher frequency variations are indicative of knots or other defects within the wood of the log. The knot signal is separated from the wood signal, for example, by applying a high pass filter to the signals corresponding to a longitudinal axis of the log. The filter produces a map in which high frequency changes in density are highlighted, and correspond to the longitudinal locations of knots. The trajectories of the knots are computed by performing a line parameter space (Houg

Affiliated with

Also associated with

Rating

Multi-view x-ray imaging of logs does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Multi-view x-ray imaging of logs, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multi-view x-ray imaging of logs will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3309025