D. Earl Kline

Computer Vision Hardware System for Automating Rough Mills of Furniture Plants

The rough mill of a hardwood furniture or fixture plant is the place where dried lumber is cut into the rough parts that will be used in the rest of the manufacturing process. Approximately a third of the cost of operating the rough mill is the cost of the raw material. Hence any increase in the number of rough parts produced from a given volume of raw material can markedly affect profit margins of a company. To automate this initial cutup requires a computer vision system that can locate and identify surface defects on boards. This paper describes continuing research aimed at developing such a vision system. An important part of this research activity is the design effort going into creating a prototype hardware system, a system that will be able to scan variable width, variable length hardwood boards at industrial speeds of two to three linear feet per second. This system is being designed to handle full length boards up to sixteen feet long. The components of the prototype are a materials handling system, an imaging system, a image processing hardware system, and a software system for performing the necessary recognition tasks and for performing all the necessary control functions. The design of each of these components will be described with the emphasis placed on hardware development.

Automated hardwood lumber grading utilizing a multiple sensor machine vision technology

Over the last 10 years, scientists at the Thomas M. Brooks Forest Products Center, the Bradley Department of Electrical and Computer Engineering, and the USDA Forest Service have been working on lumber scanning systems that can accurately locate and identify defects in hardwood lumber. Current R&D efforts are targeted toward developing automated lumber grading technologies. The objective of this work is to evaluate hardwood lumber grading accuracy based on current state-of-the-art multiple sensor scanning technology, which uses laser profile detectors, color cameras, and an X-ray scanner. Eighty-nine red oak boards were scanned and graded using Virginia Techs multiple sensor scanning system. The same boards were also manually graded on a normal production line. Precise board grade was determined by manually digitizing the boards for actual board defects. A certified National Hardwood Lumber Association (NHLA) employed lumber inspector then graded the lumber to establish a certified market value of the lumber. The lumber grading system was found to be 63% accurate in classifying board grade on a board-by-board basis. While this accuracy may seem low, the automated lumber grading system was found to be 31% more accurate than the line graders, which were found to be 48% accurate. Further, the automated lumber grading system estimated lumber value to within less than 6% of the NHLA certified value, whereas the line grader overestimated the lumber value by close to 20%. Most automated lumber grading discrepancies resulted from board geometry related issues (e.g. board crook, surface measure rounding, calculation of cutting units, etc.). Concerning the multiple sensor scanning system, defect recognition improvements should focus on better methods to differentiate surface discoloration from critical grading defects. These results will help guide the development of future scanning hardware and image processing software to more accurately identify lumber grading features.

Using Value Stream Mapping to Analyze an Upholstery Furniture Engineering Process

This study presents a systematic approach of streamlining an upholstery furniture engineering process based on a case study in one of the largest export-oriented furniture manufacturers in China. T...

Improving on Shadow Price Information for Identifying Critical Farm Machinery

A method is presented for identifying critical farm machinery in a linear programming context. The method uses a technical coefficient sensitivity analysis formula that overcomes problems associated with direct use of shadow prices for critical machinery identification. Case studies show the formula identifies the benefits of altering machinery generally with less than 10% error.