North Carolina State University
The Science of Papermaking and Paper Recycling: A Research Experience for Teachers
Lab Manual:
Manual Image Analysis of Paper for
Dirt Counting
Author:
Richard Venditti
Associate Professor
Department of Wood and Paper Science
Raleigh NC 27695-8005
Email:
Manual Image Analysis of Paper for Dirt Counting:
Objective: Be able to perform a manual dirt count of paper.
Background: A lack of cleanliness is the number one reason that paper is rejected for quality issues. It is therefore very important to understand how dirt counting is performed. Two major methods exist: manual and automatic dirt counting. This lab will show you how to perform a manual dirt count to evaluate paper. This is useful for teaching applications where an automatic dirt counter is not available.
Automatic dirt counting equipment is very prevalent in the paper industry. The basic steps are as follows. A sheet of paper is scanned with a flat bed scanner resulting in an image. The smallest elements of the image, pixels, are each assigned a grey scale value, where 0 is black and 255 is white. The computer then processes the image making a decision whether or not each pixel is dark dirt or the bright background. This is done by having the operator input a grey scale threshold which indicates that any pixel less than the threshold should be considered dirt. Through its program, the computer groups adjacent dark pixels into objects and calculates the size of the dirt particle. Finally, the computer determines the fraction of the paper that is covered by these recognized dirt spots.
Procedure:
1. Measure the area of a sheet of paper that you would like to perform a dirt count on in units of mm2. For very dirty sheets use a smaller area, for very clean sheets use a larger area. You should be counting at least 10 spots for a reasonable analysis.
2. Circle with a pencil all of the dirt spots you can detect.
3. Match each dirt spot with the spot of closest size on the provided dirt chart. After matching the spot, somehow mark the spot to avoid double counting it.
4. Record in a tabular form the number of dirt spots of each size you have found.
5. Multiply the number of spots of each size times the size in mm2 and sum all of these areas to give you the total area covered by dirt.
6. The cleanliness is most often reported as a parts per million (PPM). A value of X PPM means that there is X mm2 of dirt on 1,000,000 mm2 of paper. Calculate the PPM as 1,000,000 * dirt area/analyzed area.
ALTERNATE PROCEDURE IF PARTICLES ARE TOO SMALL TO COUNT
Often, ink particles are sub-visible and can not be counted. These particles, if black, cause the paper to appear to be grey. If this is the case, it may be useful to simply determine a grey scale value of a paper sample. A grey scale value can be specified in a simple way by matching the greyness of a paper sample with the grey bars attached at the end of this lab. The scale shown gives white as the grey scale value of 245 and black as 0. The grey bars can be produced using the Microsoft Word program or any other computer programming with basic drawing capabilities. In Microsoft Word, simply draw a box and then specify the interior of the box to have a “fill effect” of shading.
As an example of the utility of this method, consider the deinking of a newsprint furnish that is either floated or washed. By comparing the grey scale value of the floated or washing accepts, one can distinguish if one of the processes were more effective than the other.
Additional Activities
Calculate the average dirt size of analyzed sample.
Plot the histogram of dirt sizes (i.e., plot the number of occurrences of a dirt size vs. the size using a bar graph.
Determine whether a recycling operation preferentially removes a certain size particle by analyzing the feed, accepts and rejects of the operation and using the average dirt size and histogram of dirt sizes.
Do image analysis on the same type of copy paper or tissue on different sheets of paper keeping the results separate. The dirt appears in the sheets randomly. Does the area you chose to analyze and your technique give a reproducible result? Try to improve on the reproducibility by changing the size of the area analyzed.
Qualitative Grey Scale Identifier
White
245
240
235
230
225
220
215
210
205
200
195
190
185
180
175
170
165
160
155
150
145
140
135
130
125
120
120
115
110
105
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
Black