Skip to main content
ARS Home » Plains Area » Lubbock, Texas » Cropping Systems Research Laboratory » Cotton Production and Processing Research » Research » Publications at this Location » Publication #67699
Title: UPDATE ON THE DEVELOPMENT OF A BARK AND GRASS INDICATOR FOR COTTON GINS

Author
item Barker, Gary
item KECK, ROBERT - LOCKEED MARTIN
item Byler, Richard

Submitted to: National Cotton Council Beltwide Cotton Conference
Publication Type: Proceedings
Publication Acceptance Date: 1/11/1996
Publication Date: N/A
Citation: N/A

Interpretive Summary: The need to produce a better quality product for sale to the cotton textile mills and to reduce labor costs during processing has led to considerable interest in process control for cotton gins. One of the major problems facing producers and ginners in the stripper harvested area is the presence of bark in the ginned lint. Software for detecting bark in cotton gins was developed using an existing video color/trash meter to complement existing process control software. Simple algorithms were developed with detect bark 90% of the time when the operator classes the bark, in the viewing area, as large (greater than 0.5 inches). The algorithms utilize variable thresholding techniques and retain no knowledge of previous scans which detect trash particles, thus, multiple pieces of bark may be reported when only 1 piece exists. It is believed that these algorithms can provide valuable information to the cotton ginner during the ginning process.

Technical Abstract: This study was initiated to provide the ginner with knowledge of the presence or absence of bark (or grass) in the ginned lint during the ginning process. An existing color/trash meter was connected to a PC in a laboratory environment. Simple "run length" algorithms were developed along with variable threshold techniques, to eliminate the effects of variable lighting and to detect bark. The algorithms will detect bark 90% of the time when the piece of bark is at least one inch in length. The programs had difficulty detecting small pieces of bark (0.2 to 0.5 in). As expected, the programs also had problems distinguishing between large leaf trash and bark. The programs are suitable for incorporation into an existing process control system to aid the ginner in his decision making process. However, the programs are probably not accurate enough for use in cotton classing offices.