Enhancement of the computer lumber grading program to support polygonal defects

نویسندگان

  • P. Klinkhachorn
  • R. Kothari
  • D. Yost
  • P. Araman
چکیده

Computer grading of hardwood lumber promises to avoid regrading of the same lumber because of disagreements between the buyer and the seller. However, the first generation of computer programs for hardwood lumber grading simplify the process by modeling defects on the board as rectangles. This speeds up the grading process but can inadvertently put a board into a lower grade because rectangular approximation can cause clear board surface area to be lost. This paper presents a polygonal computer lumber grading program that models the defects on the boards as polygons, thereby making the grade determination process more accurate. The grading program presented herein allows species-independent rules and considers both faces of the board in the grading process. The approximation of defects as polygons does result in an increased amount of processing time, which can be readily offset with the fast and inexpensive computers that are now available. It is not uncommon for the same lumber to be graded several times. Part of this regrading effort arises because of disputes between the buyer and seller in interpreting the National Hardwood Lumber Association (NHLA) grading rules. Computer grading of hardwood lumber thus becomes an attractive proposition. Computers are consistent and fast, and computer-based grading is feasible more than ever due to National Hardwood Lumber Association. 1990. Rules for the measurement and inspection of hardwood and cypress lumber. NHLA, Memphis, Term. Hallock, H. and L. Galiger. 1971. Grading hardwood lumber by computer. Res. Pap. FPL-157. USDA Forest Serv., Forest Prod. Lab., Madison, Wis. Huang S. 1987. Expert systems for grading hardwood lumber. Ph.D diss. Purdue Univ., West Lafayette, Ind. Klinkhachorn P., J.P. Franklin, C.W. McMillin, R.W. Conners, and H.A. Huber. 1988. Automated computer grading of hardwood lumber. Forest Prod. J. 38(3):67-69. the ever-decreasing cost and increasing power of successive generations of computers. Computer grading of hardwood lumber has been accomplished by Hallock and Galiger. However, deficiencies with this program impede widespread commercial applications. These deficiencies include the inability to adapt the program code into a working environment with other programs to evaluate and control lumber processing and the inability to extend the grading to specialized species. Probably its greatest limitation lies in the consideration of only one face in the grading process. These deficiencies have been addressed by Huang and Klinkhachorn et al. Both these efforts consider both faces in the grading process and can be extended to incorporate species-dependent rules. These efforts, however, still suffer from a disadvantage. The disadvantage arises from the rectangular modeling of the defects. Defects such as wane and knots are seldom rectangular in shape (Fig. 1). A rectangular modeling of the defects on a sample board, as shown in Figure 2, eliminates a substantial amount of clear wood from being considered in the grading process. Rectangular modeling of defects is not without its merit. It reduces the amount of time required by the computer in determining the allowable clear cutting areas on the board, and hence, the grade of a board. The authors are, respectively, Associate Professor and Research Assistant Professor, Dept. of Electrical and Computer Engineering, West Virginia Univ., Morgantown, WV 26506; Development Engineer, Syncro Development Inc., 580 Middletown Blvd. Suite D203, Langhorne, PA 19047; and Project Leader, USDA Forest Serv., Brooks Forest Products Center, Virginia Tech., Blacksburg, VA 24061. The authors wish to acknowledge the USDA Forest Serv., Southeastern Forest Expt. Sta. for partial financial support. © Forest Products Research Society 1992. Forest Prod. J. 42(10):41-46. FOREST PRODUCTS JOURNAL Vol. 42, No. 10 41 Figure 1. — A sample board. Figure 2. — Rectangular representation of the sample board. The approximation of defects as polygons does result in an increased amount of processing time, which can be offset with the faster and less expensive computers that are now available. This paper presents an enhanced computer grading program (hereafter referred to as the polygonal grading program) based on the one developed by Klinkhachorn et al. (hereafter referred to as the rectangular grading program). It is anticipated that such a polygonal modeling of defects will lead to a better estimate of the grade of hardwood boards. The polygonal grading program The polygonal grading program models the defects present on the board as convex polygons. A convex polygon is a polygon in which every point on a line segment joining two points within the boundary of the polygon also lie within the polygon. Figures 3a and 3b illustrate a convex and a non-convex polygon, respectively. Those defects that are not convex polygons are transformed into convex polygons automatically by Figure 3. — Two types of polygons are possible: a) convex; and b) non-convex. Non-convex polygons are transformed to be convex (c). This is done to ensure proper program processing of each defect, while not removing clear wood area from consideration. Figure 4 — Polygonal representation of defects more closely resembles defect boundaries and generally increases clear wood for processing. the program (Fig. 3c). This modeling considerably reduces the number of computations that are required to determine the allowable cutting areas for a given grade. Figure 4 shows the sample board shown in Figure 1 as modeled by this polygonal computer grading program. A comparison of Figure 2 (rectangular approximation) with Figure 4 (polygonal approximation) shows that the polygonal program allows more clear wood to be used in determining the grade of the board. The grading process itself is carried out much the same way as a human grader would. The physical properties of the board, such as surface measure, standard length, width etc., are first evaluated. Once these measures are ascertained, the program considers each of the grades, one grade at a time beginning with the highest obtainable grade. If no grade-reducing defects are found for the grade under consideration, the evaluation continues to find the board area available in allowable cuts for the grade under consideration. To this end, the program is divided into three major components for grading FAS, Selects, and Common grade boards according to the NHLA grading rules. Once the physical properties are evaluated, the program determines if any overlength exists. If so, the

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

LUMBER GRADING Spatial WITH A COMPUTER

Over the past few years significant progress has been made in developing a computer vision system for locating and identifying defects on surfaced hardwood lumber. Unfortunately, until September of 1988 little research had gone into developing methods for analyzing rough lumber. This task is arguably more complex than the analysis of surfaced lumber. The prime motivation for developing such a s...

متن کامل

Automated Grading of Rough Hardwood Lumber

Any automatic hardwood grading system must have two components. The first of these is a computer vision system for locating and identifying defects on rough lumher. The second is a system for automatically grading boards based on the output of the computer vision system. This paper presents research results aimed at developing the first of these components. The challenge is to create robust met...

متن کامل

Development of a 3d Log Sawing Optimization System for Small Sawmills in Central Appalachia, Us

A 3D log sawing optimization system was developed to perform log generation, opening face determination, sawing simulation, and lumber grading using 3Dmodeling techniques. Heuristic and dynamic programming algorithms were used to determine opening face and grade sawing optimization. Positions and shapes of internal log defects were predicted using a model developed by the USDA Forest Service. L...

متن کامل

Machine Vision Systems for Processing Hardwood Lumber and Logs

Machine vision and automated processing systems are under development at Virginia Tech University with support and cooperation from the USDA Forest Service. Our goals are to help U.S. hardwood producers automate, reduce costs, increase product volume and value recovery, and market higher value, more accurately graded and described products. Any vision system is composed of two bread tasks: imag...

متن کامل

Preliminary Full-Scale Tests of the Center for Automated Processing of Hardwoods’ Auto-Image Detection and Computer- Based Grading and Cutup System

Automated lumber grading and yield optimization using computer controlled saws will be plausible for hardwoods if and when lumber scanning systems can reliably identify all defects by type. Existing computer programs could then be used to grade the lumber, identify the best cut-up solution, and control the sawing machines. The potential value of a scanning grading system depends on the accuracy...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1997