Agriculture Reference
In-Depth Information
data showed that fiber from the coupled lint cleaner had less mill waste and produced
stronger yarn. The nep counts, however, in the raw fiber were significantly higher for
the coupled lint cleaner (Price and Gillum, 1998; Hughs et al., 1990).
A computerized process control system for cotton gins was developed at the
USDA-ARS Cotton Ginning Research Unit at Stoneville, Mississippi. The system
included sensors to measure cotton moisture, color, and foreign matter at three sta-
tions in the gin system. The sensor measurements were used to control the ginning
process with a dynamic programming model and control hardware. The model deter-
mined the optimum gin machinery sequence based on the cotton market price and
the gin performance characteristics measured by the sensors. According to decisions
made by the model, control hardware is actuated to bypass or select any combination
of four seed-cotton cleaners, two multipath driers, and three lint cleaners (Anthony,
1990). Results of this research have been used in development of the
IntelliGin sys-
tem
by Uster Technologies. The IntelliGin system has been integrated with drying
devices and cleaning machines for process control. Using various sensors, the sys-
tem measures color grade, trash, and moisture content of the cotton online at three
locations in the ginning process stream. Based on the fiber properties measured,
and for optimal fiber quality and profit, the system can automatically adjust dryer
temperature and select the number of seed cotton and lint cleaners required in the
ginning process. The IntelliGin system has been commercially available since 1998
and installed in a number of commercial gins (Williams and Jones, 1997).
In order to automatically select an optimal configuration for the cleaning process
in gins, research has been conducted to develop systems for real-time measurement
and categorization of trash during ginning. The USDA-ARS Southwestern Cotton
Ginning Research Laboratory in Las Cruces, New Mexico, has developed a machine
vision-based system to identify and categorize non-lint materials in ginned cotton.
The system uses a scanner for acquisition of cotton images and software for image
processing. The performance of the system was evaluated in comparison with AFIS
and HVI. Trash, dust, and the total counts measured by the system correlated well
with measurements by AFIS and HVI Trash Meter (Siddaiah et al., 2006). Pelletier
et al. (2000) developed a system for seed cotton trash measurement, consisting of a
color video imaging sensor for image acquisition and a computer for image process-
ing. Trash in seed cotton was estimated through analysis of an image with a special-
ized algorithm. With this technology, images of seed cotton from various cleaning
stages can be taken in real time without any sample preparation or specialized pre-
sentation of the seed cotton.
6.5 SUMMARY
Cotton is one of the world's most important agricultural crops. Many recent improve-
ments in production and processing efficiencies have centered around automation in
the form of mechanization, sensors, and controls. PA technologies for cotton have
included yield monitoring and sensing and control systems for VRA of, for exam-
ple, nitrogen. Electronic and sensing systems have also been developed for map-
ping fiber quality and profitability. RS has been used to measure crop variability,
and automated systems have been developed for commercial application of RS for
