Classification of segregation level in steel wires by image processing

This article describes a visual measurement system intended for an automatic determination of segregation level in steel wires. The suggested system is principally based on image processing methods for segmentation and classification of inhomogeneous regions inside an otherwise uniform wire. Until our automated inspection system has been deployed, the measurement of segregation level was performed by an inexact and subjective evaluation by a human inspector. The visual measuring system involves an image acquisition stage, image pre-processing, wires segmentation and finally wires classification. In the following text all these parts of evaluation process are described in-depth. Finally each wire is automatically classified into the one of several predefined classes by the mentioned sequence of image processing steps.