ConductVision Materials Science

Grade pilling from a photo

Photograph a specimen after your pilling test, and ConductVision suggests its grade from 1 to 5, marks the areas that look most pilled and gives the probability of each grade.

A person on your team checks each suggestion and decides the grade you report.

Close view of the ribbed knit cuff of a dark jacket, with small pills and loose fibers on its surface
Pills on the ribbed cuff of a worn jacket.
  • Grades from 1 to 5
  • The probability of each grade
  • A heat map on each photo
  • The same grade each time a photo is graded

Visual assessment

Graders do not always agree

At the end of a pilling test, a grader compares each specimen by eye with reference fabrics or photographs and reports a grade.1,2

When two graders are a grade or more apart, ISO's grading standard has them grade the specimen again. If they still differ, the report gives the lowest and highest grades as well as the average.3

ASTM calls the between-laboratory precision of its random tumble test poor. It does not recommend that test for acceptance testing, and advises caution if a lab uses it that way.1

ConductVision grades every photo the same way. Its heat map marks the areas that look most pilled, so you can check a grade against its photo later.

ISO pilling grades
  1. Orange knit fabric with a smooth surface5 No change
  2. Blue knit fabric with a few loose fibers on its surface4 Partially formed pills
  3. Pink knit fabric dotted with small pills3 Moderate pilling
  4. Gray knit fabric with tangled fibers and large pills across it2 Distinct pilling
  5. Gray knit fabric covered in pills1 Severe pilling

At grade 3, pills of varying size and density cover part of the specimen. At grade 2 they cover a large proportion of it, and at grade 1 all of it.3

Each photo shows a cotton blend fabric at the grade given by graders at the Zhejiang Institute of Quality Sciences.

Who it is for

Teams that grade pilling

Mill and knitter quality labs

Graders on every shift get a second opinion from the same software version, with a heat map to check against the specimen.

Brand and retail quality teams

Check the grades your suppliers report against ConductVision's grade for a photo of each specimen.

Test labs

When a client questions a result, you can show the photo, its heat map and the version that graded it.

Fiber, yarn and finish R&D

Photograph every sample of a new yarn or finish under the same setup, and you can compare grades taken weeks apart.

What you get

What each result shows

A suggested grade

One whole grade from 1 to 5 for each photo.

The probability of each grade

ConductVision's probability for each grade, so you can spot a borderline specimen.

A heat map

A map laid over your photo marking the areas that look most pilled.

The software version

Each result names the ConductVision version that graded it, and that version gives the same photo the same grade every time.

How it fits

One photo after the test

ConductVision grades a photo you take after the test, so you run the test as you do now.

  1. Run your pilling test

    ConductVision is set up for Martindale specimens tested to ISO 12945-2. If your lab runs another test, start by sending photos of specimens you have already graded.

  2. Photograph each specimen the same way

    Keep the light, angle and distance the same for every photo, so differences between photos come from the fabric.

  3. Add the photos to ConductVision

    Each photo comes back with a grade, the probability of each grade and a heat map.

  4. Decide the grade you report

    Your grader looks at the suggestion, the heat map and the specimen, then sets the grade.

Pilling tests

Pilling tests and what ConductVision grades

Each test wears the fabric in a set way, and then a grader assesses the surface by eye. Of the tests below, ConductVision is set up for the Martindale test in ISO 12945-2.

  • Martindale

    The tester rubs a round specimen against the same fabric, or against wool fabric where relevant, in a Lissajous figure under a set force. A grader assesses it after set stages of rubbing.4

    ISO 12945-2, ASTM D4970

  • Pilling box

    Specimens on polyurethane tubes tumble at random in a cork-lined box turning at a constant speed. A grader assesses them after set stages.5

    ISO 12945-1

  • Random tumble

    Specimens tumble at random in a lined cylinder for a set time before a grader assesses them. ISO's version excludes fabrics that cannot tumble freely, and ASTM's is meant for woven and knitted apparel fabrics.1,6

    ISO 12945-3, ASTM D3512

  • Brush pilling

    The fabric is worn on a brush pilling tester. ASTM intends this test mainly for upholstery, automotive, luggage and heavy-duty uniform fabrics, because it is highly abrasive.7

    ASTM D3511

Visual assessment under ISO 12945-4

What counts as a pill

A pill is a ball of tangled fibers standing up from the fabric, dense enough that light does not pass through it and it casts a shadow. Pills can form in washing, dry cleaning and wear.3

Pilling, fuzzing and matting

The standard grades pilling, fuzzing and matting separately, on most woven and knitted fabrics, including fleeces. ConductVision's grade is the pilling grade.3

Half grades

When a specimen looks between two grades, the grader reports a half grade, such as 3-4. ConductVision returns whole grades, so your grader decides any half grade. The probability of each grade shows when a specimen is borderline.3

Viewing conditions

Graders view specimens in a darkened room, under daylight-type light set at 5° to 15° to the specimen, from 30 to 50 cm away. The standard notes that a different viewing setup can lead to different results.3

Try it on your fabrics

What to send

  • Photos or scans of tested specimens, covering the grades you usually see
  • The grade your lab gave each specimen
  • The test method, with the rubs, cycles or run time you used
  • What each fabric is, such as a cotton knit or a polyester woven

With your lab's grades, we can show how often ConductVision's grade matches your graders' grades on your own fabrics.

Other textile measurements

Dimensional stability & skewWarp and weft shrinkage plus skew, measured from marked-square photographs before and after wash.Cover factor & open areaOptical open area, pore size distribution, and cover uniformity measured on the actual fabric.Foreign fiber contaminationFind polypropylene, coloured, and foreign fibers in the web before they reach the dyehouse.Fiber diameter distributionMean diameter, CV%, and D10/D50/D90 for wool, synthetic, and nanofiber, measured from the image.

Send photos of specimens you have graded

Include the grade your lab gave each photo. Our team will grade the photos with ConductVision so you can compare its grades with yours.

Photo credits