Steel Strip Surface Defect Identification using Multiresolution Binarized Image Features

  • PDF / 1,346,371 Bytes
  • 11 Pages / 595.276 x 790.866 pts Page_size
  • 52 Downloads / 219 Views

DOWNLOAD

REPORT


TECHNICAL ARTICLE—PEER-REVIEWED

Steel Strip Surface Defect Identification using Multiresolution Binarized Image Features Zoheir Mentouri . Abdelkrim Moussaoui . Djalil Boudjehem . Hakim Doghmane

Submitted: 2 March 2020 / in revised form: 22 July 2020 / Accepted: 3 September 2020  ASM International 2020

Abstract The shaped steel strip, in the hot rolling process, may exhibit some surface flaws. Their origin could be the internal discontinuities in the input product or the thermomechanical transformation of the material, during the shaping process. Such defects are of a random occurrence and may lead to costly rework operations or to a downgrading of the final product. So, they should be detected and identified as soon as possible, to allow a timely decision-making. For such a quality monitoring, the used vision systems are mainly based on an image description and a reliable classification. In this paper, we explore pre-defined image filters and work on a procedure to extract a discriminant image feature, while realizing the best trade-off between the improved recognition rate of the surface defects and the computing time. The proposed method is a multiresolution approach, based on the Binarized Statistical Image Features method, employed to date in biometrics. The filters, pre-learnt from natural images, are applied to steel defect images as a new surface structure indicator. They provide a quite discriminating

Z. Mentouri (&) Research Centre in Industrial Technologies CRTI, P.O. Box 64, 16014 Che´raga, Algiers, Algeria e-mail: [email protected] Z. Mentouri  D. Boudjehem Laboratory of Advanced Control-LABCAV, Universite´ 8 Mai 1945 Guelma, BP 401, 24000 Guelma, Algeria A. Moussaoui Laboratory of Electrical Engineering-LGEG, Universite´ 8 Mai 1945 Guelma, BP 401, 24000 Guelma, Algeria H. Doghmane Laboratory of Inverse Problems-PI:MIS, Universite´ 8 Mai 1945 Guelma, BP 401, 24000 Guelma, Algeria

image description. A relevant data reduction is used together with a classifier to allow an efficient recognition rate of the defective hot rolled products. Keywords Computer vision  Statistical features  Classification  Strip surface defects  Hot rolling process

Introduction Computer vision systems are widely used in steel surface inspection. The general concept retained is that the defect images are processed in a way that allows the extraction of the most discriminant features, to ease the defect categorization. The published works in this field, mention numerous approaches. For instance, in an earlier study, Ref. [1] used a cost matrix method to optimize features selection and to facilitate the choice of a classification method of steel flat products, in cold rolling process. Depending on the type of defects and the specificity of applications, different feature extractors have been applied, later. As reviewed by [2], morphological operations, joint spatial/frequency domain and spatial domain filtering have been found interesting for all types of steel surfaces. Indeed, the true position of micro-defects