COIR FIBRE LENGTH ANALYSIS USING IMAGE PROCESSING TECHNIQUES

Abstract

Coir fibre is traded internationally for its value in various industries. The fibre length is an important parameter of characterizing the fibres taken as raw material in many industries with a specific length for their production purposes.

This dissertation presents an innovative way of analysing length distribution of coconut fibres using image-processing techniques. Further the project explores the potential of image processing, which used in realizing such applications.

In the past several years there has been remarkable concern in image processing and analysis techniques for the industrial and manufacturing field. Difficulties occur mainly due to time constraints imposed by traditional manual measurements, where the fibres are measured individually. Image processing techniques such as image enhancement, segmentation and skeletonization are utilized on scanned images of the fibres for subsequent analysis. Due to the complexity of the various possible orientations of the fibres in the sample, a specific algorithm is required for proper analysis of the length of fibres having diverse shapes.

The system implementation had been carried out using MATLAB® version 6.5.1, which is a technical computing language that supports image processing techniques and analysis.

The results achieved confirm that the system could be used successfully to obtain the length distribution of coir fibre samples with a quick and accurate mode relative to the traditional manual measurements within some limitations.