Using Image Analysis to Measure Fibre Diameter Distributions
This paper is Part 14 in a series titled Fundamental Principles of Fibre Diameter Measurement.
Abstract
Image analysis is today and enormous field, due largely to the rapid advances in computer technology. Perhaps the most spectacular applications, simply because the impact on the general community, are in the many areas of modern medical tomography, such as Magnetic Resonance Imaging (MRI), Ultra Sound Imaging and Computerised Axial Tomographic (CAT) scans.
Another significant and well known application is in the X-ray imaging equipment now commonplace in security applications such as screening luggage and airports.
Image analysis is used extensively by NASA for researching the many bodies in our universe and routinely for analysing satellite photographs for estimating agricultural production around the world of a wide range of crops.
Image analysis can be broadly defined as deriving useful information from images. In many applications these images are constructed from complex electromagnetic spectral data using high speed computers, rather than directly from photographic images.
In the case of wool fibre diameter image analysis is generally defined as deriving an estimate of fibre diameter by examining a photographic, digital or projected image of the fibres, either manually, electronically or digitally.
The earlier article in this series on the Projection Microscope categorised this technology as Optical Microscopy. The Laserscan instrument was placed under the category of Photometry. However both of these technologies incorporate elements of Image Analysis.
In the case of the Projection Microscope, the measurement is derived making manual measurements of fibres from a magnified image projected onto a screen. During the early development, images of fibre cross sections were also photographed and subsequently measured. Projection Microscopy can therefore also be categorised as the analysis of an image.
The Laserscan also uses image analysis, in the optical detector and in the discriminator. The optical detector analyses a diffraction pattern projected onto the detector by measuring a drop in the voltage output of the detector. The discriminator is a crude digital detector consisting of 17 “detectors” – the electronic circuitry uses the signal output from these detectors to decide whether or not a fibre image projected on to the discriminator meets the criteria established to determine whether or not the ‘image” cast onto the detector is to be measured. In a similar vein, the Electronic Fibre Fineness Indicator (EFFI), developed by Sheep and Fur Animal Branch, Animal Husbandry Research Division, ARS, USDA in Beltsville Maryland in 1957 also uses elements of Image Analysis.
In presenting the story of the different technologies explored by the wool industry in its search for more accurate and more efficient technologies for measuring this most important characteristic of wool the classification of both of these technologies under another heading always was a matter of convenience. Such is poetic, or should we say, scientific licence. However, it does mean that in outlining the development of image analysis as a tool for estimating fibre distribution characteristics of wool, we can start the story in comparatively recent times.
This article traces the development of image analysis technologies for measuring wool fibre diameter distributions. It discusses early attempts to automate the Projection Microscope, the use of television by the Micro Image Analyser, the Fibre Diameter Video Analyser (FIDIVAN), the Fibre Image Display and Measurement instrument (FIDAM) and finally the Optical Fibre Diameter Analyser (FIDAM).
Citation
Fundamental Principles of Fibre Diameter Measurement. Part 14: Image Analysis, P.J. Sommerville, AWTA Ltd Newsletter, April 2007