Faculty of Mechanical Engineering Brno University of Technology 1/4 Review of Doctoral Thesis 1. PhD candidate Ing. Jakub Hurník / Jakub.Hurnik@vut.cz 2. Name of PhD programme Design and Process Engineering (Mechanical Engineering Design) 3. Title of PhD thesis Measurement of Shape and Dimensions of Forgings 4. Principal supervisor doc. Ing. Daniel Koutný, Ph.D. / Daniel.Koutny@vut.cz 5. Co-supervisor Ing. Aneta Zatočilová, Ph.D. / zatocilova.a@fme.vutbr.cz 6. Reviewer doc. dr. Drago Bračun/ Drago.Bracun@fs.uni-lj.si University of Ljubljana 7. Overview of the scope of PhD thesis1 Very good Precise open die forging is very important to achieve the correct shape of forgings and thus reduce the amount of post-processing, reduce scrap and increase the quality of the product. This dissertation focuses on the development of a silhouette-based measurement system for dimensional control of large glowing forgings. Laser-based measurement systems typically lead the way in this type of measurement due to their accuracy and insensitivity to environmental effects. However, their disadvantage lies in the safety aspects and high price. The ideal solution for dimensional control would be to measure only with cameras that are spatially distributed throughout the forge and capture images of the forgings from a safe distance. The development of such a system, sufficiently accurate and insensitive to disturbances from the environment, is a great challenge that has not yet been properly solved. The main objective of this work is to investigate the characteristics and limitations of the silhouette-based measurement method. In this regard, the work has several sub-objectives with the main goal of reliable camera calibration, accurate edge detection and measurement to validation in an industrial environment. The work includes a literature review, a summary of the missing knowledge, the objectives of the work, scientific questions and hypotheses, explanation of materials and methods, and results and discussion. In addition, there are four scientific papers in which the candidate explains the main achievements, from the improvement of the robustness of the circular coded target system for camera calibration, the system calibration performed in parallel with each measurement, 1 Overview of the scope of PhD thesis is a short description of objectives of PhD thesis’s research and summary of main findings and scientific achievements. Faculty of Mechanical Engineering Brno University of Technology 2/4 improving immunity to interferences from the forge workshop, to the validation of the system in an industrial environment. 8. Significance of the topic and clarity of problem statement Very good The measurement of large forgings is of great importance to forging industry, as the shape and dimensional inaccuracies of heavy forgings must be constantly monitored to prevent deviations and to correct them when necessary. This includes measuring the length, diameter, and especially the straightness of the axis. The problem is clearly defined. In short, to investigate the characteristics and limitations of the measurement approach based on object silhouettes in the industrial environment when measuring heavy forgings. In this respect, the work has four objectives. First, to investigate the application of an error correction method for circularly encoded targets used in camera calibration for industrial environments. The goal is to achieve reliable camera calibration. Second, to develop an online camera calibration method based on resection of space to prevent dimensional instability. Third, to research an accurate edge detection method for measuring forging silhouettes, and fourth, to develop a forging silhouette measurement system and test its precision in an industrial environment. In addition, there are two scientific questions (1) How to implement error correction method in cyclic codes decoding of circular coded targets used for camera calibration? and (2) How to suppress corrupted edges during silhouette measurement? For each of these questions, the hypothesis and reasons are stated. I evaluate that the problem statement is clearly stated. 9. Knowledge of existing literature Very good The thesis includes a literature review with 63 references from different fields, from the review of measurement methods for measuring glowing forgings to the different algorithms used in the process. I evaluate that the literature review is sufficient and adequate. 10. Choice of methods and technical soundness Very good The thesis focuses on the goal of improving silhouette-based measurement method through research, development, and systematic evaluation. The study systematically evaluates existing active and passive optical measuring methods and proposes the original solutions from the system conceptual design, image processing to system calibration and validation using experimental data collected in laboratory and industrial environments, with particular attention to the effects of the industrial environment. I would rate the technical soundness as high since the work uses appropriate methods, techniques, tools and addresses potential sources of bias and error. 11. Quality, originality and significance of the results Very good An important component of the silhouette-based measurement method is the multi camera calibration and its verification in an industrial environment. To measure the dimensions of forgings, the distance of the cameras to the forging and their mutual position must be determined. Unlike the laboratory environment, the industrial environment is unstable because of subsequent changes due to vibration of heavy equipment, high temperatures, and dust, and the calibration is not stable over time. The originality of the presented method lies in the calibration, which is performed in parallel with each measurement. For this purpose, Faculty of Mechanical Engineering Brno University of Technology 3/4 coded patterns are used, which are placed near the forging so that they are visible in all images taken from different angles. For this method to work, the exact distances between the coded patterns must be known in advance, and most importantly, the coded patterns must be made of such materials and placed in such a way that they survive near the forging, where there is strong thermal radiation, a lot of dust, and the possibility of mechanical damage, which is disadvantage of the proposed method. For this purpose, the research paid special attention to improving the robustness of the circular coded target system for camera calibration. In addition, for accurate measurements, the original image processing was developed to detect the edges of the forgings in the images. The research also confirms the influence of an irregular forging cross-section on the determination of the diameter and improvements when using multiple cameras from different angles. In summary, significant original results have been obtained with great potential for application in the industrial environment. 12. Quality of attached papers Excellent The candidate has published research in three peer-reviewed journals with scientific citation factor: (1) Circular coded target system for industrial applications, Machine Vision and Applications (IF 2.983, 2021); (2) Enhancing the accuracy of forging measurement using silhouettes in images, Measurement (IF 5.131, 2021); (3) Multi‑view camera system for measurement of heavy forgings, The International Journal of Advanced Manufacturing Technology (IF 3.563, 2021); He also presented his work at scientific conference: Camera calibration method of optical system for large field measurement of hot forgings in heavy industry, Proc. SPIE 11056, Optical Measurement Systems for Industrial Inspection XI, (2019). I consider the publications and journals mentioned to be of high quality. 13. Overall assessment, strengths and weaknesses (based upon the above evaluation categories 8–12) Very good The dissertation deals with the improvement of the silhouette-based measurement method for dimensional control of large glowing forgings, which is of great importance to the forging industry because the shape and dimensional inaccuracies of large forgings need to be constantly monitored to avoid deviations and expensive low quality. The study systematically evaluates existing active and passive optical measurement methods and proposes original solutions, paying particular attention to the disturbances of the industrial environment. The method, as conceived, is very practical for use in the workshop. Install spatially distributed cameras in the forge, take pictures of the forgings from a safe distance and calculate something from the images. However, the problem lies in the vibration of the heavy equipment, high temperatures and dust, and consequently the calibration is not stable over time. The originality of the presented research lies in the calibration, which is performed in parallel with each measurement. For this purpose, coded patterns are used, which are placed near the forging so that they are visible in all images taken from different viewpoints. For this method to work, the exact distances between the coded patterns must be known in advance, and most importantly, the coded patterns should be able to withstand strong thermal radiation, a lot of dust, and also mechanical damage, which is a drawback of the method. The candidate is aware of this problem and pays special attention to improving the robustness of the circular coded target system for camera calibration and proves that his method can achieve acceptable accuracy. To this end, several other details are also improved, such as the original image processing to detect the edges of the forgings in the images, or the influence of an irregular forging cross-section on the determination of the diameter. The candidate has published three research papers in peer-reviewed journals with scientific citation factor and at a SPIE conference, which is also an indicator of the high quality of the research work that brings the method of passive forging measurement closer to implementation in the industrial environment. Faculty of Mechanical Engineering Brno University of Technology 4/4 14. Questions and comments 15. Conclusion PhD thesis is an independent scientific work that presents a novel solution to a significant problem in the research area and demonstrates the candidate’s ability to conduct independent research. YES 16. Date and signature 20/02/2023 Please note A. Evaluate categories 7 to 13 using the following scale: unacceptable, acceptable, satisfactory, good, very good, excellent. The qualification of ‘excellent’ should only be given for a PhD Thesis in the top 3% of the research in your field of expertise. B. E-mail the completed form to: Klara.Javorcekova@vut.cz