Seite - 122 - in Proceedings of the OAGM&ARW Joint Workshop - Vision, Automation and Robotics

Automated Quality Assessment of Remelted Steel Ingots Daniel Gruber1, Harald Ganster1 and Robert Tanzer2 Abstract—For high quality steel products it is essential to have specific understanding of the underlying steel production process such as the electric slag remelting process (ESR). To assist the currently manual assessment there is a high need for objective quality measures and standardized evaluation methods. A set of relevant parameters can be derived from the so-called pool profiles that give insight to the remelting process. Based on texture segmentation and ridge detection a computer-vision based automated evaluation of the pool profiles is achieved. A comparison with manually extracted pool profiles from expert metallurgists shows the feasibility of the approach and the good performance of the automated analysis. Further evaluation on different types of steel blocks will yield valuable insight to and improve the overall steel production process. I. INTRODUCTION AND MOTIVATION The field of quality management and improvement in high quality steel production is one of the deciding reasons whether a steel producer remains competitive or not. In the production of high quality steel products for demanding applications it is essential to remelt conventional produced ingots. In order to yield specific understanding of the remelt- ing process as well as to improve the process, there is a high need foran objective andstandardizedevaluation of remelted blocks. The advantage through technology is to be able to sub- stitute pure manual quality control and, thus, very time- consuming work flows. Furthermore, it is possible to provide repeatable calculations of quantitative measurements. This paper presents a vision-based solution to be able to automate those processes. Currently, most of the structure evaluation is done man- ually and the information is stored in different analog and digital files. In order to be able to store all information in one place, a software was developed where various different kinds of meta data can be directly mapped to the analyzed steel block. After a short introduction of the data material (Section II) and a brief overview of related work (Section III), Section IV gives insight into the quality assessment of steel ingots. Section V presents the automated segmentation, Section VI an objective method to derive steel quality parameters and Section VII gives some final conclusions and an outlook on future work. 1JOANNEUM RESEARCH Forschungsgesellschaft mbH, DIGITAL - Institute for Information and Communication Technologies, Austria daniel.gruber@joanneum.at, harald.ganster@joanneum.at 2BO¨HLER Edelstahl GmbH & Co KG, Austria robert.tanzer@bohler-edelstahl.at II. DATA MATERIAL Figure 1 shows the scheme of an ESR. Those remelted high quality steel blocks have a weight up to 20 tons. To analyze the inner solidification of such blocks, it is necessary to saw out longitudinal slices from the center of the block. Furthermore, these plates are cut into pieces to be able to handle size and weight as illustrated in Figure 2. Fig. 1: Scheme of electric slag remelting process. Fig. 2: Preparation of steel blocks for evaluation. Inorder togaindeeperknowledgeof the remeltingprocess those plates are ground, polished and etched to reveal the inner crystalline solidification structure. Those structures provide information directly linked to the remelting param- eters and as a consequence are essential for optimizing these parameters. Changes within the remelting process are directly related to the solidification structure [11], [8]. As a 122