Introduction to SeedCount X

8/8/2019 Introduction to SeedCount X

1/8

What are SeedCount and CornCount?

SeedCount and CornCount are digital imaging systems specifically designed for the grain

industry. They use software and flatbed scanner technology to rapidly and accurately analyse a

sample of grain and determine its physical characteristics. They generates detailed data tables

that can be exported to any spreadsheet or database program and provide this detailed

information without damaging the grain sample.

SeedCount and/or CornCount are for use by anyone who grows, sells, buys or uses grain and/or

assesses its quality. Potential users are breeders, brewers, dealers, growers, food processors,

maltsters, millers, etc.

SeedCount

SeedCount consists of a reflectance scanner and is currently offered for barley, wheat, rice and

corn analysis. The following table shows the grain type and parameters measured using

SeedCount.

Wheat Barley Corn(Maise) Rice

Ave Kernel Area Ave Kernel Area Ave Kernel Area Ave Kernel Area

Ave Kernel Length Ave Kernel Length Ave Kernel Length Ave Kernel Length

Ave Kernel Width Ave Kernel Width Ave Kernel Width Ave Kernel Width

Seed Count Seed Count Seed Count Seed Count

KKW KKW KKW KKW

Dockage % Dockage % Dockage Dockage

Plumpness Plumpness Red Streak Chalkiness

Blacktip Blackpoint Crowns Whiteness

Dents Discoloured SeedsStress Cracks

Horneous

Endosperm

CornCount

CornCount consists of two scanners, ie, an incident (reflectance) based scanner (i.e., a

SeedCount scanner) and a transmission based scanner (Epsom V500 scanner). A corn tray is


2/8

used in a reflectance mode to measure size, ie, length, width, area, crowns, dents and colour. A

transmission tray is used to detect stress cracks and Horneous Endosperm.

Future versions will include additional trays and calibrations for other grains and cereals.

How Does SeedCount Work?

SeedCount uses a modified flatbed desktop scanner, a sample tray and a Microsoft Windows

based personal computer to create a digital image of a sample of grain, and then

analyse the image. The scanner operates facing down inside the instrument cabinet.

See Figure 1.

A sub-sample of the grain is obtained with a sampling tube that collects a sample "core" from

all levels in a bucket or small bag of grain. The tube contents are transferred to a volumetriccup. Up to 718 barley, 1000 wheat, 1300 rice or 316 corn kernels can be analyzed at once.


3/8

Sampling Tube Sampling Tube

The grain cup is filled with grain and levelled off with a score plate. The sample of grain is then

poured onto the special indented tray,and shaken to distribute the seeds into the shaped

indents. The indents have varying shapes allowing some seeds to fall into wide, shallow indents

and lie on their back. In this position the length, width and area of the seeds can be measured.

Other seeds fall into narrower indents and are held on their edge. When on their edge their

thickness can be measured. Some grains, e.g. corn, also use end-on indents for viewing the

crown directly.

The tray is placed into the scanner cabinet which scans the sample at 300 dpi

in 16.8 million colours. The image can be saved to the computers hard drive as a

lossless JPEG image file or as a BMP file for future reference.


4/8

The user enter the clean weight of the sample and can also enter the initial as-is

weight, moisture percentage, protein percentage and the volume of the sample. The

user and site identification can also be entered. The operator then presses the Analyse button

and the software computes the various parameters for the specific grains or seeds. Figure 3.

shows the main computer screen with the results of the Analysis routine.

As the grain is not damaged by the scanning process, it can be retained for retesting or

used for other purposes.

The program can save all of this data in a convenient form for use in a spreadsheet or database.

The results of the Analysis routine can be used to generate a selection of plots.


5/8


6/8

SeedCount Image Analysis of Grains and SeedsImage analysis is an opto-mechanical measurement system to determine the pixel by pixel

characteristics of a seed or grain or any other product or item. The basic parameters that image

analysis measures are:

Edge Detection: defines the edge of an item by contrasting the colour of the background to the

colour of the item.

Length: defines the horizontal dimension of an item by identifying the pixels at the edges of the

item and then multiplying the number of pixels by the resolution of the scanner.

Width: defines the vertical dimensions of an item by identifying the highest and lowest pixels at

the edges of the item and then multiplying the number of pixels by the resolution of the

scanner.

Colour: defines the Red, Green, Blue, shades of each pixel in the image. The R, G, B, values can

be expressed as alternate colour coordinates, ie, L*, a*, b* or X,Y,Z.

Grey Scale: defines the black to white shading of the object.

SeedCount uses specially designed sample trays to position seeds in a 3 dimensional fashion.

Two rows of indents in a tray, narrow and wide, are used to allow the seeds to lie flat or on

their edge. Images of the seeds in the tray are used to measure characteristics of seeds and

grains. The following measurements are available within the SeedCount software:


7/8

Seed Count: the number of whole seeds detected in the image. Seeds that are lying in contact

with another seed are not counted as whole seeds. Broken seeds and foreign matter is not

counted.

Average Kernel Length: the numerical average length of the seeds that are counted in the Wide

slots in the tray.

Average Kernel Width: the numerical average width of the seeds that are counted in the Wide

slots in the tray.

Average Kernel Thickness: the numerical average width of the seeds that are counted in the

Narrow slots in the tray.

Average Kernel Area: the area in sq. mm of whole seeds in the Wide section of the tray.

Aspect Ratio: the ratio of the width to the length.

Roundness: the average of the ratios of the length/width, length/thickness, width/thickness,

i.e.,

Roundness = (Length/Width + Length/Thickness + Width/Thickness)/3

Screening Equivalents: calculates the thickness of each seed and determines the percentage of

seeds that fit within specific ranges, egg, 0-2mm, 2.0-2.2mm, 2.2-2.5mm etc. Screenings is

expressed as percentage of the total weight of the sample.

Screening Distribution: the percent mass corresponding to slotted screen sieving

measurements.

Dockage: calculates the total dockage as the difference between the total weight of the sample

and the clean weight of the sample, plus the amount of the sample that is detected as broken

grains, awns and some weed and foreign seeds. Dockage is expressed as a percentage of the

total weight of the seeds.

Average Kernel Weight: the total weight of the sample divided by the seed count. Only includes

whole kernels.

KKW (Thousand Kernel Weight): the calculated weight of 1000 kernels based on the average

weight of the scanned seeds, expressed in grams. Can also be expressed as the average kernel

weight in grams. Both As-Is and Dry Basis can be calculated using the weight of the sample anda moisture value for the sample.

Test Weight: the calculated weight of 100 litres of grain. Expressed as kilograms per hectoliter

or pounds per bushel.

Discoloured Kernels: the percentage of Seeds that are either;

dark or mouldy

red or marked with red

green


8/8

or yellow

Blackpoint or Blacktip: the number and percentage of seeds that have a black tip on the tip of

the endosperm side of the seed. Blackpoint is related to wheat and Blacktip is related to barley.

Blackpoint Impact: the blackness and area of discolouration is assessed as the Blackpoint

Impact and indicates the severity of the defect.

Kernel Brightness: the average brightness of the seeds in the sample.

Chalk: the number and percentage of rice kernels that have white spots on the seed. A score

for each seed tested, to computes the Chalkiness of the seeds.

Chalk Impact: the amount of chalk assessed on the basis of the whiteness and the area of the

chalk.

White Seeds: the number and percentage of seeds that have exposed endosperm or fusarium

infection.

CIE L*a*b*: the average L, a, b values and standard deviation for the sample.

Crown: the percentage of corn kernels with rounded heads vs. flat heads.

Horneous Endosperm: the percentage of corn kernels area.

Dents: the Square mm of corn kernel with an indent on the flat side of the corn kernel.

Red Streaks: the percentage of corn kernels area with red streaks visible on the kernels.

Kernel Classifications: for medium, long grain, arborio rice. Some of the Rice specific

classifications include: Paddy Immature (green) grains Red seeds Red streaked seeds Red specked seeds Yellow seeds Black seeds Black specked seeds Discoloured seeds

Individual Seed Information: SeedCount scans the sample tray and displays the complete

image of the tray. When the image has been analysed using the SeedCount software, the data

for each seed is available by clicking onto the individual seeds. Each seed is identified by a

specific number and all the measured parameters are available on the screen. The image and

the results can be stored in memory.

Documents

Introduction to SeedCount X