After milling, the next operation of the HRY measurement
consists of accurately separating head rice from broken
kernels. Typically, this procedure is done by a mechanical
sizing method or by hand picking. Both of these separating
methods are operator dependent with operator subjectivity
becoming a factor, especially in hand picking. Mechanical
separating can utilize sieves, riddles, and indent cylinders.
The most common laboratory method used is a device
consisting of inclined indent plates developed by Smith
(1955). However, this device has been shown to provide
inaccurate and highly variable separation of milled rice
kernels as noted by Stermer and Beerwinkle (1970). They
compared 10 different sizers (also called shaker tables) and
noted that factors such as stability of the supporting table and
variations in vibration of the carriage table affected sizer
performance. They concluded that acceptable precision
among sizers was obtained by adjusting vibration frequency
H
644 APPLIED ENGINEERING IN AGRICULTURE
of the carriage table using a suitable variable–speed drive.
The shaker table has also been noted as time consuming,
requiring operator supervision to ensure adequate separation.
Previous researchers used other devices to separate the
different fractions of long–grain rice. Mathewson et al.
(1990) built a mechanical system to separate a milled rice
sample into four fractions using sieves. The system required
30 to 40 min to separate 100 g of milled rice into fractions.
They evaluated their sieving system using an image analysis
system, placing five subsets of 45 to 60 kernels (or broken
kernels) on a black velvet background. By capturing the
projected area of the kernels, each piece was categorized as
whole kernels or brokens.