Pseudo Generation of Metallographic Images and Verification of Superiority for Discrimination Problems -Applying Adversarial Generative Networks-

Metallography is the study of the structure of metals and alloys. Metallographic analysis can be regarded as a detection tool to assist in identifying a metal or alloy, to evaluate whether an alloy is processed correctly, to inspect multiple phases within a material, to locate and characterize imperfections such as voids or impurities, or to find the damaged areas of metallographic images. However, the defect detection of metallography is evaluated by human experts, and its automatic identification is still a challenge in almost every real solution. Deep learning has been applied to different problems in computer vision since the proposal of AlexNet in 2012. In this study, we propose a novel convolutional neural network architecture for metallographic analysis based on a modified residual neural network (ResNet). Multi-scale ResNet (M-ResNet), the modified method, improves efficiency by utilizing multi-scale operations for the accurate detection of objects of various sizes, especially small objects. The experimental results show that the proposed method yields an accuracy of 85.7% (mAP) in recognition performance, which is higher than existing methods. As a consequence, we propose a novel system for automatic defect detection as an application for metallographic analysis.

Download Full-text

Problems of Analyzing Microstructure Images in Assessing the Impact of Technological Parameters of Combined Strain Wave Hardening on the Quality of the Surface Layer

10.30987/graphicon-2019-2-261-264 ◽

2019 ◽

Author(s):

Андрей Киричек ◽

Andrey Kirichek ◽

Дмитрий Соловьев ◽

Dmitriy Solovyev ◽

Александр Хандожко ◽

...

Keyword(s):

Surface Layer ◽

Thermal Treatment ◽

Strain Wave ◽

Technological Parameters ◽

Steel Microstructure ◽

Main Indicator ◽

Metallographic Images ◽

Surface Layer Quality ◽

The Impact

The problems of analyzing metallographic images and the method of their solution using modern software for the analysis of metallographic images are described. There is given an analysis of microstructure images as the main indicator of the surface layer quality by the example of studying the research results of strain wave hardening combinations and chemical-thermal treatment, in particular the influence of previous strain wave hardening and subsequent thermal and chemical- thermal treatment on the alloy steel microstructure or previous thermal and chemical- thermal treatment and subsequent strain wave hardening. On the basis of the analysis the effectiveness of strain wave hardening and chemical and thermal treatment is established.

Download Full-text

Algorithm for Automatic Pattern Classification Designed for Real Metallographic Images

Advances in Multimedia and Interactive Technologies - Video Surveillance Techniques and Technologies ◽

10.4018/978-1-4666-4896-8.ch013 ◽

2014 ◽

pp. 165-179

Keyword(s):

Pattern Classification ◽

Threshold Value ◽

Mean Value ◽

Process Quality ◽

Steel Plant ◽

Central Axis ◽

Grey Scale ◽

The Mean ◽

Metallographic Images

The problem of automatic pattern classification in real metallographic images from the steel plant ArcelorMittal Ostrava is addressed. The goal is to monitor the process quality in the steel plant. In the images of metal, there are dark dots that are produced by imperfections along the central axis of each plate. It is necessary to determine automatically the number and sizes of these dots. The number and sizes of the dots is a measure of how imperfect each plate is. The process is presented that segments the area of plates that contains segregation, identifies those rows of pixels along which the dots lie, and counts the pixels that are marked as dots by evaluating all the vertical columns of pixels that intersect the rows that contain the dots. The threshold value is set to be 95% of the mean value of grey scale for each column of pixels and makes the dots white. White dots that are most likely noise are removed to identify dots that are smaller than 4 connected pixels across. The explanations related to the obtained results are firmly related to the information provided by human experts.

Download Full-text

Family of models for describing one class of metallographic images

Pattern Recognition and Image Analysis ◽

10.1134/s1054661809020187 ◽

2009 ◽

Vol 19 (2) ◽

pp. 334-341 ◽

Cited By ~ 6

Author(s):

D. A. Perfil’ev ◽

G. M. Tsibul’skii ◽

Yu. A. Maglinets

Keyword(s):

Metallographic Images

Download Full-text

Automatic Pattern Classification of Real Metallographic Images

2009 IEEE Industry Applications Society Annual Meeting ◽

10.1109/ias.2009.5324864 ◽

2009 ◽

Cited By ~ 5

Author(s):

Vesna Zeljkovic ◽

Pavel Praks ◽

Robert Vincelette ◽

Claude Tameze ◽

Ladislav Valek

Keyword(s):

Pattern Classification ◽

Metallographic Images

Download Full-text

Using the Principles of Image Analysis in the Assessment of the Proportion of Retained Austenite in the Case Hardened Layers

Archives of Metallurgy and Materials ◽

10.1515/amm-2017-0085 ◽

2017 ◽

Vol 62 (2) ◽

pp. 577-580

Author(s):

P. Váňová ◽

J. Sojka ◽

T. Kulová ◽

K. Jokešová ◽

P. Purtátor ◽

...

Keyword(s):

Image Analysis ◽

Retained Austenite ◽

Volume Fraction ◽

Hardened Layer ◽

Carbon Plate ◽

Surface Saturation ◽

Metallographic Images ◽

Dimensional Instability ◽

Spontaneous Conversion

Abstract Structures of carburized layers after the surface saturation process in gaseous, liquid or solid medium and after subsequent heat treatment (hardening and low-temperature tempering) consist mainly of high carbon plate martensite with a certain portion of retained austenite. The presence of retained austenite (RA) in carburized layers is mostly considered as undesirable because it decreases hardness of the hardened layer and furthermore, a spontaneous conversion to a ferritic-carbide mixture of a bainitic type, accompanied by a change of properties, dimensional instability and the local increase in internal stress with the possible formation of cracks, can occur. The proportion of retained austenite is, therefore, a significant characteristics of the quality of hardened layers. This work deals with the evaluation of the volume fraction of retained austenite in carburized layers using image analysis on metallographic images.

Download Full-text

Influência da atmosfera de sinterização na densificação de uma liga Monel 400

Revista Liberato ◽

10.31514/rliberato.2019v20n34.p169 ◽

2019 ◽

pp. 169-176

Author(s):

Angelo Ores Bonamigo ◽

Jorge Luis Braz Medeiros ◽

Luciano Volcanoglo Biehl ◽

Hektor Oliveira Borges ◽

José de Souza

Keyword(s):

Solid Medium ◽

Compaction Pressure ◽

Inert Atmosphere ◽

The Other ◽

Argon Gas ◽

Gaseous Environment ◽

Monel 400 ◽

Metallographic Images ◽

Gas Medium ◽

Porous Area

Resumo O presente estudo compara a densificação de amostras de uma liga de níquel Monel 400 sinterizadas em duas diferentes atmosferas, sendo uma atmosfera inerte de gás argônio e outra um sólido composto de coque de petróleo. Três pares de amostra foram feitos, cada uma delas compactadas em diferentes pressões de 665 MPa, 1350 MPa e 2000 MPa; nos pares de mesma pressão, uma amostra foi sinterizada no meio gasoso, e a outra amostra sinterizada no meio sólido a uma temperatura de 1100°C por um tempo de 180 minutos. Diversas imagens metalográficas foram obtidas e analisadas por meio de software, para descobrir a porcentagem de área porosa superficial. Verificou-se que as amostras que tiveram maior pressão de compactação obtiveram uma menor área porosa, assim como as sinterizadas, em meio sólido, também obtiveram uma maior densificação, em relação às amostras sinterizadas em meio gasoso. Palavras-chave: Densificação. Sinterização. Atmosferas de sinterização. Abstract The present study compares the densification of samples of a sintered Monel 400 nickel alloy in two different atmospheres, one being an inert atmosphere of argon gas and other a solid medium composed of petroleum coke. Three sample pairs were made, each compressed at different pressures of 665 MPa, 1350 MPa and 2000 MPa; in the same pressure pairs, one sample was sintered in the gaseous environment, and the other sample sintered in the solid environment at a temperature of 1100 ° C for a time of 180 minutes. Several metallographic images were obtained and analyzed by software to find the percentage of porous surface area. It was found that the samples that had higher compaction pressure had a smaller porous area, as well as the sintered ones, in solid medium, also had a higher densification, compared to the sintered samples in gas medium. Keywords: Densification. Sintering. Sintering atmospheres.

Download Full-text

Electropolishing of Stainless Steel in Laboratory and Industrial Scale

Metals ◽

10.3390/met9080854 ◽

2019 ◽

Vol 9 (8) ◽

pp. 854 ◽

Cited By ~ 2

Author(s):

Paweł Lochyński ◽

Sylwia Charazińska ◽

Edyta Łyczkowska-Widłak ◽

Andrzej Sikora

Keyword(s):

Stainless Steel ◽

Photoelectron Spectroscopy ◽

Passive Layer ◽

X Ray ◽

Force Microscopy ◽

Atomic Force ◽

Process Scale ◽

Metallographic Images ◽

Industrial Samples ◽

Difficult Issue

Transposing the process scale from laboratory to industrial conditions is a difficult issue that applies to many sectors of the industry. As far as electropolishing of stainless steel is concerned, the limitations connected with a significant increase in the area of electropolished surface should be considered, along with the possibility of defects that may emerge. This paper compares the results of electropolishing of stainless steel in the laboratory and in industrial conditions. For the analyzed conditions, it was determined that the best results, both in laboratory and industrial conditions, were obtained at temperature of 35 °C and current density of 8 A·dm−2. High temperatures resulted in the emergence of defects on the surface, in particular for industrial samples. The defects were visualized by metallographic images with Nomarski contrast and atomic force microscopy. X-ray photoelectron spectroscopy tests were used to analyze the composition of the passive layer on the electropolished surfaces.

Download Full-text