Guide s21 r472 Ing

GU

I

D

E

S21 Realese 4.72

LKL Tecnologia da Informao Ltda. - Avenida Batista Botelho, 604 - Centro

Santa Cruz do Rio Pardo - SP - Brazil - ZIP Code 18900-000

Tel: +55 14 3372 6702 - E-mail: [email protected] - Internet: www.lkl.com.br

L K L T e c n o l o g i arice statistical analyzer

Presentation

S21 Classier - Guide

The following instructions have the goal of addressing the functionality of the S21 classier

software, aspects of calibration and system maintenance.

Firstly, the user will know the capabilities of the system through access to seven pre-scanned

samples and the application of some available lters.

The user may also, in addition to creating new lters, change the parameters of the existing

lters and understand the logic of function of them through the results that will be presented

from the 7 pre-scanned samples. All the percentages and numerical results has its visual

correspondent on the screen.

Starting the Software "S21 Classier"

01) To start, click on the icon "Classicador S21"

02) Insert the login (s21) and password (If you do not have a password, please request to our

support team)

s21

****

pg. 02





03) Wellcome to S21 Classier!!

04) On this screen the user must choose the sort settings available according to the class of rice

to be analyzed. To begin our explanation, choose "Long White" and click "Choose".

pg. 03





05) This is the main screen of S21. The software was designed to be simple to operate. The

basic parsing operation boils down to: a) scanning the sample; b) processing images; c)

apply a lter classication; d) view the results report.

06) All scanned and processed sample is stored in the S21. To access it click "Processed

Samples".

pg. 04





07) On this screen, enter the period you want to list the samples already processed.

ATTENTION!! To access the pre-scaned samples, you must report on the rst eld a date

less than or equal to for that the samples be listed. Then click "Load Samples". 30/09/2013

The seven samples will be listed.

30/09/2013

08) The 7 pre-scanned samples will be listed. On this screen, the user should search and

choose the sample that want to open. Continue selecting the sample "Diversos - Long White

Thin - Harvest Rice". Then click "View".

pg. 05





09) The chosen sample (Long White Thin - Harvest Rice) was opened. By default, when we open

a processed sample, the screen that shows the whole grain of the sample is displayed. At

this stage of the analysis, we have known the following results: integers, 3/4, 1/2 and 1/4

grain. Other results will come with the use of lters.

10) To access other screens that show the grain 3/4, 1/2 and 1/4 of the sample click one of the

options of the "Image" menu.

pg. 06

Sorting

Sorting





Percentage of 3/4 in the sample "Diversos - Long White Thin - Harvest Rice" = 9:54%

Percentage of 1/2 in the sample "Diversos - Long White Thin - Harvest Rice" = 3.09%

Percentage of 1/4 in the sample "Diversos - Long White Thin - Harvest Rice" = 0.01%

pg. 07

Sorting

Sorting

Sorting





11) To continue to explain the use and functioning of lters, turn back to the whole grain screen.

Click the item "Whole" of the "Image" menu.

12) The lters in S21 are logical structures that apply to any processed sample to nd the

indexes of industrial quality of the sample. The results are used to calibrate the industry,

keeping a standard of rice quality, from the arrival till the nal product. In research centers,

the lters allow to evaluate the physical characteristics of rice. The S21 works with two

groups of lters: "Basic Filters" and "Advanced lters". To access them, a sample must be

open. We will begin by explaining Basic Filters, but before accessing them read the box on

the next page.

Basic FilterBasic FilterBasic FilterAdvanced FiltersAdvanced FiltersAdvanced Filters

pg. 08

Sorting

Sorting





Before starting the use of lters, it is necessary to understand how the S21 processes the

samples.

After processing the images coming from the scan, the S21 individualizes the grains found in the

images with the following physical characteristics:

Length, Width, Chalk Area, Spotted Area, Total Whiteness and Vitreous Whiteness.

Click on a frame (grain) of the screen and see the individual characteristics of the selected grain.

Individual Characteristics of Grains

The S21 analizes one grain at a time. When there are two or more grains in the same frame,

always consider the grain at the center. See the example below.

1

pg. 09





Basic Filters

The Basic lters are simple logical structures and easy to use. Its ltering logic is simple and

gives the S21 user a quality prole of the sample that ts perfectly in the industrial classication

process of rice.

To continue the explanation, we will use the same sample as an example: "Diversos - Long White

Thin - Harvest Rice".

01) Click Basic Filter.

02) This screen shows the list of basic

lters available. Has four basic lters

created, but you can create as many

lters as needed.

To apply or edit a lter, select an item

from the list. In this example, as we

have a sample of Long rice type,

choose the lter "Long White Rice".

pg. 10

Sorting





03) The box will expand showing the saved parameters in the lter. These parameters dene

with what criteria (chalk, spoted, length and width) that rice will be analyzed. The box at the

bottom of the page has information about these parameters. Now we proceed with the

analysis. Click "Apply

This parameter of the Basic

Filter denes the index of

chalky grains in two levels:

Chalky (grains with chalk

area >= 70%) and

Partially Chalky (grains

with chalk area >= 50%

and < 70%). These

parameters are editable.

This parameter of the Basic

Filter denes the index of

the grains with Color

defects (will be considered

grains with color defects

those with spotted area

>= 10.5%). This

parameter is editable.

Filter Long White Rice. How it Works?

This parameter of the Basic Filter, denes (in this example) the index

of grains of Other Varieties that are not Long (will be considered

"other varieties", grains with width between 2.5 mm and 3 mm). This

parameter is also editable.

Obs.: The checkbox "Active" has the function to enable or not the

parameter of the lter.

pg. 11





04) After apply the chosen lter, the S21 visually presents the results of the applied lter by

marking the grains with the colors of the indexes.

From this stage of the analysis, the user can print the report with the results.

However, before printing the report, will be showed some important additional features of

the Basic Filter.

05) To improve visualization of the results, we can group the grains within their indexes. Click

the item "Whole" of the "Damage" menu.

pg. 12

Sorting

Sorting





06) Now the grains screen is displayed with the grains grouped within its index. To see the

gruped grains click the indices shown on the top screen.

Chalky 0,58 %

Partial Chalky 3,81 %

pg. 13

Sorting

Sorting

Sorting





Color Damage 0,54 %

Other Varieties 0,00 %

Sorting + Drag and Drop Grains

Another feature of the Basic Filter, very useful in the classication of the sample, is the Sorting

of grains according each of its features and the use of the mouse to move the grains with the

"Drag and Drop Grains" feature.

Through this feature you can merge two types of analysis: Numeric (only the application of the

lter) and Visual (sensory). The advantage of this hybrid process in relation to the conventional

analysis (manual) is the quickness with which the analysis is done by S21.

To show this feature we'll use as an example the "Partial Chalky" index that is in 3.81%. We will

manipulate the grains visually by using the feature "Sorting + Drag and Drop grains".

The partial chalk is a index with a high degree of subjectivity, it is impossible to visually

dierentiate, for example, a grain with 50% of chalk area of another that has 48% or 53% of

chalk area. The human eye does not have this capability.

Therefore, starting from the beginning of this subjectivity, we can move some grains visually and

suit the results.

pg. 14

Sorting

Sorting





See how to use the feature "Sorting + Drag and Drop Grains" in the S21:

01) Click the index "Partial Chalky" to see the grains found by the lter. The index is at 3.81%.

Remember that we are manipulating the sample "Long White Thin - Harvest Rice".

02) Now click the index "Head Rice" to see the good grains. The index found by the lter is

82.43%

pg. 15

Sorting

Sorting





03) The goal now is to nd within the "Head Rice", Partial Chalky grains. To quickly nd them,

we will use the Sorting feature. Click the "Sorting" menu and then click the "Chalky" item.

With this command, we sorted the grains in the screen Head Rice from more chalk to less

chalk. Probably the rst grain on the top screen will have a little less than 50% of chalk area

and decreasing in the sequence. These rst grains were not considered "Partial Chalk" by

being with the chalk area just below the value set by the lter, in this case the limit was set

at 50%.

With the sort feature from the S21, the search for grains with more chalky within the "Head

Rice" screen will be fast and simple.

Now the user can use the Drag and Drop Grains feature of the S21 to put the grains visually

deemed partial chalk into the index "Partial Chalky".

See how on the next page.

descending order of chalky area

pg. 16

Sorting





01) Hold the "Ctrl" key and click with the mouse on the grains that you judge partial chalk,

selecting them.

02) The S21 user can combine the "CTRL" key with the "SHIFT" key to select an entire

sequence of grain. Click the rst grain of the sequence holding the "CTRL" key. Release the

"CTRL" key. Now click the last grain of sequence holding the "SHIFT" key.

CTRLCTRLCTRL

SHIFTSHIFTSHIFT

pg. 17

Sorting

Sorting





03) Click the right button of the mouse on a highlighted grain. Holding the button, drag and

drop the grains within the index "Partial Chalk".

+

+

04) After dragging and dropping grains of "Head Rice" to "Partial Chalk" the index will be

increased.

Comments About Sorting + Drag and Drop Grains Feature

With regard to Sorting + Drag and Drop

Grains feature, the user can apply them in

various situations, see some examples:

a) Dismember the "Color Defects" index in

indices with specic characteristics, eg:

Yellow, Spotted, Striped, etc.

b) Find whole grains without tips.

c) Discard to the trash the impurities that are

not part of the sample.

d) Find poorly polished grains.

e) Find red grains on brown rice.

f) Analyze index of peeled rice (peeled x

paddy).

If you have questions, please contact the

support team.

pg. 18

Sorting

Sorting





Printing of Results

01) To view and print the results click "View Report".

02) The report will be displayed. Is required an installed printer driver to open the Atention!

display. To print click "Print".

pg. 19

Sorting





S21 Statistical report

User: Takeshi

Identication of the report: 201310021051

Classication setup: Long White

Date/Hour: 02/10/2013 10:51:25

Origin of Sample: Diversos

Sample identication: Long White Thin - Harvest Rice

Report setup: Long White Rice

Report setup description: Kernels with chalky area >= 70.00 % will be cons idered "Chalky". Grains with chalk area

b etween 50.00 e 70.00 are considered Partial Ch alk. Kernels with spotted area >= 10.50 % will

b e considered "Color Damage". Kernels with width between 2.500 mm and 3.000 mm will be

c onsidered "Other Varieties".

Whole average height: 5.752 mm

Whole average width: 2.071 mm

Kernels dropped to the trash: 1 Kernels

*** -

Quality Scores

Whole total: 87.37 %

62.48 gr

Processed rice: 100.00 %

71.52 gr

H ead Rice 82.43 %

58.95 gr

C olor Damage 0.54 %

0.39 gr

C halky 0.58 %

0.42 gr

P artial Chalk 3.81 %

2.73 gr

B roken 12.63 %

9.04 gr

Total whiteness: 131.60

White vitreous: 122.21

Additional Information

Wholes Three-quarters Half One-quarter Total

Kernels 1888 435 338 4 2665

% Weight 87.37 9.54 3.09 0.01

% Other Varieties 0.00 0.08 0.00 0.00 0.08

% Color Damage 0.54 0.43 0.72 0.00 1.70

% Chalky 0.58 0.21 0.05 0.00 0.85

% Partial Chalk 3.81 0.75 0.22 0.00 4.78

% Chalky area 12.64 4.04 3.76 0.03 20.46

PPC: 178

03) The reports issued by the Basic Filters follow a standard format.

The report of the Basic Filter is divided into three parts:

"Heading"

"Quality Indexes"

"Aditional Information".

pg. 20





Header

Standar Report of the Basic Filter

Quality Scores

The header section contains the data that identies the analysis ( ).A

We have in the header the information about the average length and average width of the grains

considered whole and also the grains discarded for trash ( ). The recycle bin feature B will be

discussed later, has the function to clean up and discard particles and/or grains that are

unwanted in the scanned sample.


User: Takeshi


Classication setup: Long White

Date/Hour: 02/10/2013 10:51:25


Sample identication: Long White Thin - Harvest Rice

Report setup: Long White Rice

Report setup description: Kernels with chalky area >= 70.00 % will be cons idered "Chalky". Grains with chalk area

b etween 50.00 e 70.00 are considered Partial Ch alk. Kernels with spotted area >= 10.50 % will

b e considered "Color Damage". Kernels with width between 2.500 mm and 3.000 mm will be

c onsidered "Other Varieties".




*** -

A

B

Quality Scores

Whole total: 87.37 %

62.48 gr


71.52 gr

H ead Rice 82.43 %

58.95 gr

C olor Damage 0.54 %

0.39 gr

C halky 0.58 %

0.42 gr

P artial Chalk 3.81 %

2.73 gr

B roken 12.63 %

9.04 gr



The Quality Scores section contains qualitative data of the analysis: ( ) Total of whole grains in A

percentage and weight. The total sample weight (72.52 g) is a data informed by the user. Just

below we have qualitative data presented in two columns: ( ) Percentage and ( ) weight. Lastly B C

two indices of whiteness: ( ) Total Whiteness and Vitreous Whiteness. See the next page the D

comment about the whiteness in the S21.

B C

A

D

pg. 21





Aditional Information

The "Additional Information" section contains qualitative data of the sample separated by the

columns: Whole, 3/4, 1/2 and 1/4. The last line of the session gives us a new qualitative

data, that is only possible to measure by analysis of digital image: Chalk area ( ) of the A

sample. The smaller the chalk area more vitreous is the sample.

About Total Whiteness and Vitreous Whiteness in the S21

Regarding the measurement of whiteness in

the S21, has two indices: Total Whiteness and

Vitreous Whiteness. The S21 has its own

scale of whiteness with range 0-200.

The Total Whiteness is measured from the

total mass of the sample.

The Vitreous Whiteness is measured from the

total mass of the sample discarding the part

considered chalk.

We use the Vitreous Whiteness as a polishing

reference, because it has less inuence of the

amount of chalk present in the sample.

To get an idea, whiteness 40 in Kett scale

corresponds to 130 in the S21 scale.

In most varieties of white rice, the Vitreous

Whiteness should be between 120 and 127

after polishing.


Wholes Three-quarters Half One-quarter Total

Kernels 1888 435 338 4 2665

% Weight 87.37 9.54 3.09 0.01

% Other Varieties 0.00 0.08 0.00 0.00 0.08

% Color Damage 0.54 0.43 0.72 0.00 1.70

% Chalky 0.58 0.21 0.05 0.00 0.85

% Partial Chalk 3.81 0.75 0.22 0.00 4.78

% Chalky area 12.64 4.04 3.76 0.03 20.46

PPC: 178

A

pg. 22

30/09/2013





Advanced Filters

The Advanced Filters are more complex logic structures that enable to elaborate various types of

analyzes. To create and/or edit the advanced lters is necessary that the user has knowledge of

programming logic.

This guide does not address issues concerning elaborating and editing of advanced lters. The

default software comes with some examples of advanced lters for that the user to know the

potential of this group of lters.

To access the list of Advanced Filters, you have to open some Processed Sample.

01) Click Processed Samples.

02) Enter the period that you want to list the samples already processed. Attention!! To list the

pre-scanned samples, you must report in the rst eld the date for that the 30/09/2013

samples be listed. Then click "Load Samples".

pg. 23





03) The seven samples will be listed. On this screen the user can search and list the samples

that he want to open. , choose the sample "Diversos - Basmati Rice To continue training

from Brazil" and then click "View".

04) With the Basmati rice sample openned, access the list of Advanced Filters. Click the item

"Aplly Filter" of the "Advanced Filters" menu.

pg. 24

Sorting





05) The S21 has 8 advanced lters saved. The

lters of the class "Histogram" were created

to attend the research centers that need the

analysis of the sample with a frequency

diagram, grouped by each of the

characteristics of the grains: Chalk, Length,

Total Whiteness, Vitreous Whiteness, Dark

Spotted and Spotted.

The last lter of the list, "White Long Thin -

Brazilian Quality Normative", was created to

assist in the supervision of the long thin rice

type in Brazil, according to Brazilian

normatives of quality.

In this example, we will apply the lter

"Histogram - Length White" in the Basmati

rice sample to know in detail the length ranges

of the grains of the sample.

04) Click the "Histogram - Lenght White".

The tab with the logic tree of the lter will open, but do not bother to understand it, at this

point we just want to apply the lter in the sample of the Basmati rice. Click "Apply

pg. 25





05) After applying the chosen lter we will have the tree of the lter with the values found in

each branch of the tree. To see the result of the analysis, click "View Report".

06) To print the

report. Click

"Print".

pg. 26






User: Takeshi


Classication setup: Basmati White

Date/Hour: 02/10/2013 16:21:15


Sample identication: Basmati Rice from Brazil

Report setup: Histogram - Lenght White




*** -

Quality Scores


100.00 gr

< 1.5 0.00 %

0.00 gr

> =1.5 e < 1.75 0.00 %

0.00 gr

> = 1.75 e < 2.0 0.00 %

0.00 gr

> =2.0 e < 2.25 0.00 %

0.00 gr

> = 2.25 e < 2.5 0.00 %

0.00 gr

> = 2.5 e < 2.75 0.00 %

0.00 gr

> = 2.75 e < 3.0 0.00 %

0.00 gr

> = 3.0 e < 3.25 0.00 %

0.00 gr

> = 3.25 e < 3.5 0.03 %

0.03 gr

> = 3.5 e < 3.75 0.03 %

0.03 gr

> = 3.75 e < 4.0 0.03 %

0.03 gr

> = 4.0 e < 4.25 0.00 %

0.00 gr

> = 4.25 e < 4.5 0.03 %

0.03 gr

> =4.5 e < 4.75 0.00 %

0.00 gr

> = 4.75 e < 5.0 0.12 %

0.12 gr

> = 5.0 e < 5.25 0.08 %

0.08 gr

> = 5.25 e < 5.5 0.38 %

0.38 gr

> = 5.5 e < 5.75 1.07 %

1.07 gr

> = 5.75 e < 6.0 1.59 %

1.59 gr

> = 6.0 e < 6.25 4.87 %

4.87 gr

> = 6.25 e < 6.5 9.05 %

9.05 gr

> = 6.5 e < 6.75 14.26 %

14.26 gr

> = 6.75 e < 7.0 16.96 %

16.96 gr

> = 7.0 e < 7.25 19.01 %

19.01 gr

> = 7.25 e < 7.5 15.79 %

15.79 gr

> = 7.5 e < 7.75 8.55 %

8.55 gr

> = 7.75 e < 8.0 5.41 %

5.41 gr

> = 8.0 e < 8.25 2.11 %

2.11 gr

> =8.25 0.62 %

0.62 gr




T otal

K ernels 1 820

% Chalky area 1 9.50

PPC: 179

pg. 27





Quality Scores


100.00 gr

< 1.5 0.00 %

0.00 gr

> =1.5 e < 1.75 0.00 %

0.00 gr

> = 1.75 e < 2.0 0.00 %

0.00 gr

> =2.0 e < 2.25 0.00 %

0.00 gr

> = 2.25 e < 2.5 0.00 %

0.00 gr

> = 2.5 e < 2.75 0.00 %

0.00 gr

> = 2.75 e < 3.0 0.00 %

0.00 gr

> = 3.0 e < 3.25 0.00 %

0.00 gr

> = 3.25 e < 3.5 0.03 %

0.03 gr

> = 3.5 e < 3.75 0.03 %

0.03 gr

> = 3.75 e < 4.0 0.03 %

0.03 gr

> = 4.0 e < 4.25 0.00 %

0.00 gr

> = 4.25 e < 4.5 0.03 %

0.03 gr

> =4.5 e < 4.75 0.00 %

0.00 gr

> = 4.75 e < 5.0 0.12 %

0.12 gr

> = 5.0 e < 5.25 0.08 %

0.08 gr

> = 5.25 e < 5.5 0.38 % 0.38 gr

> = 5.5 e < 5.75 1.07 % 1.07 gr

> = 5.75 e < 6.0 1.59 % 1.59 gr

> = 6.0 e < 6.25 4.87 % 4.87 gr

> = 6.25 e < 6.5 9.05 % 9.05 gr

> = 6.5 e < 6.75 14.26 % 14.26 gr

> = 6.75 e < 7.0 16.96 % 16.96 gr

> = 7.0 e < 7.25 19.01 % 19.01 gr

> = 7.25 e < 7.5 15.79 % 15.79 gr

> = 7.5 e < 7.75 8.55 % 8.55 gr

> = 7.75 e < 8.0 5.41 % 5.41 gr

> = 8.0 e < 8.25 2.11 % 2.11 gr

> =8.25 0.62 % 0.62 gr



07) The "Quality Scores" section of the report shows how the grains are distributed by sample

length range.

08) If you installed MS Oce 2010 (32-bit) package you can export saved reports to Excel

spreadsheet. If you do not have the export module enabled, contact the support team. By

applying, for example, a column chart on the above results we get the following display:

pg. 28





09) One of the features of the S21 is that every presented result has its visual correspondence, it

was already seen in the Basic Filters. It is possible to visualize the grains that are part of the

percentage index found. To view the grains in the Advanced Filters, you must click on the

branches of the tree.

In this example below, we apply in the Basmati rice sample the advanced lter "Histogram

- Length White" to know the percentage of grains per length range. If we want to visualize

the grains (1.07%) that are in the range 5.50mm to 5.75mm we should go to the branch of

the tree that corresponds to this result and click on the branch.

10) The grains will be showed

pg. 29

Sorting





11) To return to the tree of results click the item "Filter Tree" of the "Advanced Filters" menu.

Access to the S21 Reports

01) When we apply a lter in a processed sample, we have the option to save the report with

the results. To access saved reports click View.

02) A new window will open. Through this

window the user can search and open all

the reports of saved results. There is no

saved report. Try saving a result to be

able to access it when you nd it

necessary.

pg. 30

Sorting





Trash

The S21 has a feature called the Trash, which allows elimination of undesirable particles from the

sample already scanned. It is usually used to remove impurities. Importantly, the samples must

be clean before being scanned, samples with impurities or bran will impair the results.

01) To proceed with the demonstration of the "Trash" feature, list the pre-scanned samples and

choose the sample "Long White Thin - Only Whole".

12

3

4

02) Then, to access the "Trash", click on the item "Trash" of the "Image" menu.

pg. 31

Sorting





Besides impurities, we can eliminate unwanted grains, as it will showed in this example.

Before proceeding with this example, it is important to note that in a sample, besides the broken

grains, there are two classes of whole grains:

a) Perfect Whole Grains - grains that after the hulling

and polishing remains perfect, with no broken parts.

Is also excluded from the perfect whole grain class

the smallest grains, poorly formed due to external

factors, such as lack of nutrients, moisture, light,

etc.

b) Whole Grains - are grains that are

larger or equal to 3/4 of its total.

Even if they are clipped will be

considered whole. In most

countries there are regulations

that dene the measures of

integers, 3/4, 1/2 and 1/4 for

each class of grain (long,

medium, short). Later we will

discuss how these classes work in the S21.

As we already have seen earlier in the guide, the S21 show us the sample divided into 4 size

groups: whole, 3/4, 1/2 and 1/4.

The objective of this analysis is to know what is the average length of the Perfect Whole Grains

present in the sample 'Long White Thin - Only Whole'.

We should then eliminate from the sample by dragging to the trash, the whole grains that are not

perfect. For this we use the "Trash" feature combined with the "Sorting" feature.

In this sample, the screen of "Trash" shows the number of valid grains (1322) and grains in the

trash (0).

03) Click the item 'Height' of the 'Sorting' menu.

The goal is to sort the beans in descending order by length.

pg. 32

Sorting





04) After sorting by length, scroll to the bottom of the screen. Note that there are broken and clipped

grains.

Descending order

per length

What Can Be Dragged to "Trash".

Peels:

Grains in shell: Poorly polished Grains:

Use Sorting by Spoted:

Bonded Grains in Heigth:

Bonded Grains in Width:

Use Sorting by Height:

Use Sorting by Width:

Note: The sorting is always in descending order.

pg. 33

Sorting





05) Mark the broken, clipped and poorly formed grains and with the right button of the mouse

pressed, drag and release to "Trash" .If have any questions how to mark the grains, see how to

proceed in page 14.

+

+

06) The unwanted grains were to trash. Remain valid in the sample only the perfect whole grains.

pg. 34

Sorting

Sorting





07) Selecione o item Apply lter do Menu Advanced lters.

08) Selecione o item Histogram - Lenght

White, em seguida clique em Apply.

pg. 35

Sorting





08) Click View Report

09) You can save the report.

10) The preview of the report will open. The result Average Length of Perfect Whole Grains is

in the header section of the report.

pg. 36

Sorting

1) Prepare a sample of polished rice with 40 to 100 grams. Clean the impurities and excess bran.

2) Open the S21 Classier Software.





How to scan a sample.

s21

****

3) Enter your password.

4) Choose the type of rice to be

analyzed.

5) Make sure that the lamp is lit.

4) If the receiver of sample is in the open position, put it in the closed position.

Opened Closed

To open, lift the

receiver and rotate it

counterclockwise

To close, rotate the

receiver clockwise and

down.

pg. 37





5) Dump the sample into the sample

receiver. Do not forget to put the

plastic cup at the end of the ramp.

7) Click Capture

Closed

8) Enter the sample data.

a) Source of the sample.

b) Sample Identication.

The samples are grouped by source. The source list

is editable.

This eld is used to enter the identication of the

sample.

pg. 38





c) Weight of the sample.

Enter "100" to get the percentage results or enter the

real weight of the sample so that the S21 also shows

the results on the reported weight.

9) Click Conrm to the S21 initiate capturing images.

Opened

10) Open the receiver. 11) Turn on the

vibrator to the

rice starts down

the ramp.

12) When all the rice of the sample nish

dropping the ramp, click Stop and

Process.

To open, lift the

receiver and rotate it

counterclockwise

13) Turn o the

vibrator.

pg. 39





14) At the end of processing, the screen of whole grains is showed to the user. The sample is

saved! Now just choose a lter to complete the analysis. See how to access the lters on

pages 10 (Basic Filters) and 23 (Advanced Filters).

How to Access Processed Samples.

All samples processed are saved in the

database of the S21. To access them

click Processed Samples.

To list them, inform the period and click

Load Samples. To search in the list, use

the search eld and click Find. To open,

select the desired sample, then click View.

pg. 40

Sorting





Calibration and Fine Tuning

The S21 is installed and delivered calibrated. The measurement is performed using reference

samples with known values .

There are two types of reference samples:

A) Sample with known values of whiteness and chalk area percentage.

B) Sample with known values of whole and Broken.

First, lets give attention to the concept of adjustment of Whiteness and chalk.

The S21 comes with one reference sample Whiteness and chalk.

N o t e

Fine Tuning of Whiteness and Chalk

The S21 is scheduled to ask the Fine-tuning every 60 samples or every 2 days. When prompted

we proceed with Fine Tuning by a standard sample, with known values of whiteness and chalk

area percentage.

In principle, any samples already analyzed by the S21 can be a reference sample

for calibration. When passing an old sample in the S21 it should present very

similar values . Therefore, it is necessary to keep 1 reference sample to keep the

S21 in the original calibration.

pg. 41





How to Run the Fine Tuning?

01) Clean the blue track and the white point attached on the left edge of the ramp.

02) Scan the Reference Sample.

03) Click Fining Tuning

pg. 42





04) Enter the values of the Total Whiteness Area and Total chalk area of the sample.

05) Click Set

06) The S21 will automatically adjust according to the reported values.

pg. 43





C o n c e p t s o f w h i t e n e s s a n d C h a l k i n S 2 1

Reference Sample: is a sample with the known results of Total Whiteness and Total Chalk

Area. The standard sample has a prerequisite that it be a prepared Crystalline Rice and polished

with a soft cloth so that it is free of dust stuck in its surface. This way we guarantee greater

stability in its characteristics.

Fine Tuning: Is the operability of the

S21 that adjusts the reference of

reading Whiteness and Chalk. This

reference is measured with the

processing of the standard sample.

The program checks the result of

Total Whiteness and Total Chalk Area

obtained in processing of the sample

and compares it with the "real" value

informed of the Total Whiteness and

Total Chalk Area of the Reference

Sample, then proceeds adjusting the

reading reference of Whiteness and

Chalk for the result become closer

to the "Real", keeping the system

always calibrated.

Reading Reference of Whiteness

and chalk: Is the standard

calibration reference that the S21

needs for that the results of chalk

and color damage keep within a

single standard reading. The S21 has a own pattern that comes along with the factory standard

sample, delivered along with the S21.

Shelf Life of the Standard Sample: The stored reference sample has a shelf life of 60 days,

since stored properly, away from light, heat and humidity. After 60 days you can change their

characteristics that will result in the loss of the reference values of Total Whiteness and Chalk

Area.

You should always keep some reference samples stored with an expiration date updated.

Therefore, it is necessary to replace the old samples by new standard samples. (See the next

topic)

01) Reserve a portion of approximately 60g of prepared

crystalline rice with minimum of chalky grains and

white center. The sample must to be completely

free of impurities. Clean with a soft cloth until it is

polished.

How to make new reference samples?

60g of polished crystalline rice.

Cleaned with a cloth.

pg. 44





02) Run the Fine Tuning with an old reference sample: by this way we are sure that the

system is calibrated.

03) Capture the new standard sample three times and write down the values of Total

Whiteness and Total chalk Area. Take the average of the values and round it up.

04) Store the new standard sample with the rounded values of Total Whiteness and Chalk

Area (write these values).

Adjusting PPC

Pixels Per Centimeter

The parameter that denes the measure of the grain is called PPC (pixels per centimeter) which

determines the relationship pixels per centimeter. This parameter can be changed (adjusted) by

the user.

The S21 gets the measure of length of the grain by the 2 pixels more distant from each

other (diagonal) and the width by the 2 pixels more distant from each other, perpendicular

to the length.

The PPC value is calibrated by analyzing one sample with kown values of length and width.

Each machine has its own value of "PPC". Once adjusted, it is no longer necessary

adjustment.

pg. 45





Classication Setup.

By processing the images of the grains captured by the scanner, The

S21 software divides the grains of the sample in 4 groups of size:

Whole, 3/4, 1/2 and 1/4. The result presented after processing is

dened by the conguration we choose when the software starts.

The user can create as many congurations as necessary. These

settings are associated with the class of rice we want to analyze, for

example, long, medium and short. Each conguration, besides the

name, has three parameters to be informed:

1) Maximum height (mm): this value denes what is the maximum length to be considered as

a grain in the sample. Grains

(particles) that are above this

value will not be considered

in the analysis. For example,

if the varieties of the rice

class long thin does not

exceed 8.5mm in length, we

will use a value slightly

above: 8.7mm.

2) Maximum width (mm):

the same logic should be

used for the width. If the

varieties of the rice class

long thin, do not exceed

2.4mm in length, we use a

value slightly above, for example, 2.7mm. However, it is very common that in a sample of long

thin rice has a quantity of merge of varieties, for example, medium grains, which are larger,

measuring up till 3.0mm. It is possible, by lters, to look for the wider grains. In this case we must

increment the value of width of 2.7mm to 3.0mm.

3) Reference Measure (mm): This parameter denes how will be the results of whole, 3/4, 1/2

and 1/4. Generally each country has its own rules that denes the

measures to be adopted. In Brazil, for the rice class long thin, the grain

is considered "whole" the one that is greater than or equal to 4.5mm

long, which corresponds to a value of 3/4 of Reference Measure. See

the diagram on the next page.

pg. 46

How the S21 Divides the Sample Into whole, 3/4, 1/2 and 1/4?

6.0 mm

3/4

1/2

1/4

3.0 mm

4.5 mm

1.5 mm

Wholes Kernels

1/4 Kernels

3/4 Kernels

1/2 Kernels

Reference

Measure





In this example with the

value of 6.0mm of Measure

Reference, we will have:

Will be considered whole

grains those who possess

length greater or equal to

4.5mm, that corresponds to

the value of 3/4 of 6.0mm.

Will be considered 3/4 those

grains who possess length

greater or equal to 3.0mm

and less than 1.5mm long.

Will be considered 1/2 those

grains who possess length

greater or equal to 1.5mm and less than 3.0mm long.

Will be considered 1/4 those grains who posses length less than 1.5mm long.

How to Adjust the PPC?

01) Prepare a reference sample with

approximately 50 grams, 90% whole

and 10% broken. The broken ones

should be large, coarse. Attention!

This sample should be checked

manually with a caliper.

02) Open the Classier S21 software and choose the

conguration according to the class of rice from the

reference sample.

pg. 47





03) Ensure that the value of

the Reference Measure of

the chosen conguration

is correct. When in doubt,

read the table of pages 46

and 47 about it. The value

must conform to the

criteria used in the

reference sample.

04) Scan the Reference

Sample and Process it.

05) After processing, see the percentage results of whole grains. This value should be close

to the real result, the dierence should not exceed 2%. If the dierence is greater, check

if the value of Reference Measure is correct, or contact our technical support.

pg. 48

Sorting





06) In case it shows a greater dierence than 0.5% with the actual value, we have to adjust

the PPC. Close the window of grains, and in the main window, click on Fine Tuning

07) In the "Fine Tuning" window you

can nd the value of the PPC in

the last eld at right. If necessary

reduce the percentage result of

whole grains, increase the PPC in

2-3 points.To increase the

percentage of whole grains you

must decrease the value of PPC.

After changing click "Save PPC". Close the "Fine Tuning".

08) In the main window, click Process to reprocess the sample with the new value of PPC.

09) After processing, check out the new percentage result of whole grains. If more

adjustment is necessary in PPC, repeat steps 03-06 until it is calibrated.

pg. 49

Sorting

Sorting





Repeatability of Results

The biggest feature that attests the quality of the S21 system is its high index of

repeatability of results with very low statistical deviation. This means that if we use the

S21 to scan the same sample, several times, the results are practically the same.

Note in the following assays, the amount of grain varies within the same sample, but the results

hardly vary. This is due to the processing core of the S21 be statistical. The following assays

were performed on 5 samples (Sample 001, Sample 002, Sample 003, Sample 004, Sample

005) and analyzed 3 times (a, b, c).

Sample Filter Date Avg Lenght Avg Width L/W Whit S21 Whit Vitreous Qt Kernerls

Amostra 001 a I.N.12 - V01 - BB50 11-ago-10 6,90 2,01 3,43 133,63 124,51 1971

Amostra 001 b I.N.12 - V01 - BB50 11-ago-10 6,90 2,01 3,43 133,45 124,36 1910

Amostra 001 c I.N.12 - V01 - BB50 11-ago-10 6,89 2,01 3,43 133,78 124,52 2071













Repetibilidade de Resultados

Whole Head Rice Broken Color Damage Darkened Partial Chalky Chalky Stained Other Varieties

98,00 97,57 1,95 0,05 0,00 0,34 0,04 0,05 0,00

98,12 97,63 1,84 0,00 0,00 0,49 0,00 0,04 0,00

97,98 97,22 1,99 0,07 0,00 0,61 0,00 0,05 0,06

87,20 48,42 11,11 0,85 0,00 35,81 1,91 0,97 0,92

87,45 49,18 11,13 0,93 0,02 34,71 2,14 1,14 0,76

86,91 49,00 11,30 0,81 0,00 34,61 2,29 1,30 0,71

97,81 97,81 1,85 0,04 0,00 0,20 0,04 0,06 0,00

97,62 97,59 2,08 0,04 0,00 0,16 0,00 0,11 0,00

97,61 97,29 2,14 0,00 0,00 0,22 0,04 0,00 0,31

97,65 93,62 2,24 1,92 0,00 0,98 0,35 0,81 0,08

97,63 92,82 2,25 2,10 0,13 1,05 0,37 0,72 0,56

97,79 93,12 2,19 2,04 0,14 1,25 0,22 0,53 0,51

88,82 88,02 11,18 0,04 0,00 0,72 0,00 0,04 0,00

89,23 88,21 10,77 0,00 0,00 0,90 0,00 0,04 0,08

88,70 86,95 11,30 0,00 0,00 1,39 0,00 0,08 0,28

Sample

Amostra 001 a

Amostra 001 b

Amostra 001 c

Amostra 002 a

Amostra 002 b

Amostra 002 c

Amostra 003 a

Amostra 003 b

Amostra 003 c

Amostra 004 a

Amostra 004 b

Amostra 004 c

Amostra 005 a

Amostra 005 b

Amostra 005 c

Fonte: Pirahy Alimentos - Rua Cel. Tristo de Arajo Nbrega, 1530 - So Borja - RS - Brazil - http://www.pratono.com.br/

pg. 50





Frequently Asked Questions

The analise system by digital image was developed with the aim

to simulate the industrial classication process that sorts and

separates the rice by size, by area of chalk, by color damage and

whiteness. The industrial classication does not put people to sort

and separate the rice, the classication is done by equipment that

analyze the ow until the end, by a simple, quick and objective

way. The type of analysis done by the S21 is somewhat analogous

to industrial classication process. The best way to calibrate the

industry and determine the industrial yield of raw material that

enters the industry is by digital analysis (numerical).

Sensory analysis is more subjective, it has a greater variation of

results against digital numerical analysis, it is also more time

consuming. However, the sensory analysis can reach more

complex results often needed to evaluate the rice.

What is the difference by the

digital image analysis

(numerical) of visual analysis

(sensory)?

What is the amount of rice

required that the S21 needs in

order to do analysis

Due to the process be statistical, the S21 requires approximately

1,200 grains (particles) or more, for results to stabilize with

minimal statistical deviation. In weight, the amount should be

between 30-100 grams depending on the size of the grains to be

analyzed.

How do I keep the S21

calibrated?

Passing a reference sample with known values, that is, a sample

already analyzed by the S21. The system adjusts itself from these

values . See page 41.

When passing an old sample, the S21 should shows practically

the same results.

The percentage results in the

S21 are not in weight?

Yes, the S21 calculates the percentage weight by a projection of

the grain volume x format x density conform the transparency of

the grain. This calculation is very accurate.

What is the durability of LED

lighting?

According to our supplier, the LED has a lifespan of 70.000 hours.

How much disk space is

occupied by one analysis?

Approximately 10-15 MB, including data and storing images of

each sample.

What is the average time that

the S21 uses to analyze one

sample?

To a sample of 70g, about 2 minutes. The analysis is done in four

steps: Scan the sample / Process images / Apply the lter / View

the report.

Can I combine, in the S21, the

numerical digital analysis with the

visual analysis?

Yes, through the feature: Sorting + Drag and Drop Grains. See in

the page 14.

pg. 51





The Technical Support S21 is free, permanent and exclusive to service via e-mail to the

users who have doubts about the functioning of the system and also for correction of

possible programming errors.

The Free Technical Support does not cover data loss, data files repairs and reinstallation by

misuse of the installed softwares on the computer or hardware defects.

The support through remote connection (internet): via Logmein or TeamViewer is free during

the warranty period.

It is the customer's responsibility to backup the data. The data to copy are in the directory

C:\Classificador S21.

To engage the support:

Bicio de Britto

Email: [email protected]

Takeshi Suzuki

Email: [email protected]

Cell Phone: +55 14 96282746

Skype: takeshisuzuki

LKL Tecnologia

Phone: +55 14 3372 6702

S21 Suport

Warranties

Software:

For lifetime

Hardware:

1 year from delivery date.

pg. 52

Pgina 1Pgina 2Pgina 3Pgina 4Pgina 5Pgina 6Pgina 7Pgina 8Pgina 9Pgina 10Pgina 11Pgina 12Pgina 13Pgina 14Pgina 15Pgina 16Pgina 17Pgina 18Pgina 19Pgina 20

Documents

Guide s21 r472 Ing