A Quality Control System using Texture Analysis in Metallurgy

Object detection, recognition and texture classification is an important aspect of many industrial quality control systems. In this paper, we report on a system designed for the inspection of surfaces which has a range of applications in the area of metallurgy. The approach considered is based on the application of Fractal Geometry and Fuzzy Logic for texture classification and, in this paper, focuses on the manufacture of rolled steel. The manufacture of high quality metals requires automatic surface inspection for the assessment of quality control. Quality control systems are required for several tasks such as screening defected products, monitoring the manufactures process, sorting information for different applications and product certification and grading for end customers. The system discussed in this paper was developed for the Novolipetck Iron and Still Corporation in Russia and tested with images captured at a rolling mill with metal sheets moving at speed of up to six meters per second and inspected for several defect classes. The classification method used is based on the application of a set of features which include fractal parameters such as the Lacunarity and Fractal Dimension thereby incorporating the characterisation of surface surfaces in terms of their texture. The principal issues associated with texture recognition are presented which includes fast segmentation algorithms. The self-learning procedure for designing a decision making engine using fuzzy logic and membership function theory is also presented and a new technique for the creation and extraction of information from a membership function considered. The methods discussed, and the system developed, have a range of applications in ‘machine vision’ and automatic inspection. However, in this publication, we focus on the development and implementation of a surface inspection system that can be used in a iron and steel manufacture by non-experts to the automatic recognition system operators. Keywords-Computer vision; patterns analysis; segmentation; object recognition; self-learning; fuzzy logic; image morphology.