An End-to-End Computer Vision Methodology for Quantitative Metallography

Metallography is crucial for a proper assessment of material's properties. It involves mainly the investigation of spatial distribution of grains and the occurrence and characteristics of inclusions or precipitates.This work presents an holistic artificial intelligence model for Anomaly Detection that automatically quantifies the degree of anomaly of impurities in alloys. We suggest the following examination process: (1) Deep semantic segmentation is performed on the inclusions (based on a suitable metallographic database of alloys and corresponding tags of inclusions), producing inclusions masks that are saved into a separated database. (2) Deep image inpainting is performed to fill the removed inclusions parts, resulting in 'clean' metallographic images, which contain the background of grains. (3) Grains' boundaries are marked using deep semantic segmentation (based on another metallographic database of alloys), producing boundaries that are ready for further inspection on the distribution of grains' size. (4) Deep anomaly detection and pattern recognition is performed on the inclusions masks to determine spatial, shape and area anomaly detection of the inclusions. Finally, the system recommends to an expert on areas of interests for further examination. The performance of the model is presented and analyzed based on few representative cases. Although the models presented here were developed for metallography analysis, most of them can be generalized to a wider set of problems in which anomaly detection of geometrical objects is desired. All models as well as the data-sets that were created for this work, are publicly available at https://github.com/MLography/MLography.

Quantitative characterization of the microstructure of in situ aluminum matrix composites

A quantitative assessment of the microstructure parameters is necessary for making informed decisions on the development and adjustment of technological parameters for the production of cast metal matrix composites. This study gives an estimate of the size and distribution of the reinforcing phases in the structure of in-situ Al-Mg2Si aluminum matrix composites using an automated technique for analyzing metallographic images realized in the ImageJ open-source software with developed macros. A comparison of the quantitative parameters of the microstructure of composites in different parts of the ingot is carried out. The central regions of the ingot are distinguished by higher values of the average quantity of particles per unit of the microsection surface area in comparison with the peripheral regions. The average size of the synthesized Mg2Si reinforcing particles was 16 μm and practically did not vary in different areas.

Application of Artificial Intelligence Technologies to Assess the Quality of Structures

The timeliness of the complex automated diagnostics of the metal condition for all characteristics has been substantiated. An algorithm for the automation of metallographic quality control of metals is proposed and described. It is based on the use of neural networks for recognizing images of metal microstructures and a precedent method for determining the metal grade. An approach to preliminarily process the images of metal microstructures is described. The structure of a neural network has been developed to determine the quantitative characteristics of metals. The results of the functioning of neural networks for determining the quantitative characteristics of metals are presented. The high accuracy of determining the characteristics of metals using neural networks is shown. Software has been developed for the automated recognition of images of metal microstructures, and for the determination of the metal grade. Comparative results of carrying out metallographic analysis with the developed tools are demonstrated. As a result, there is a significant reduction in the time required for analyzing metallographic images, as well as an increase in the accuracy of determining the quantitative characteristics of metals. The study of this problem is important not only in the metallurgical industry, but also in production, the maritime industry, and other engineering fields.

Processing and analysis of images of microstructure metals for determining the grain point

An algorithmic support for metallographic images preprocessing and analysis is presented. The software product implements metallographic methods for the grain size determination by comparison of rating scales, counting beans, calculation of grain boundaries intersections for equiaxed and elongated grains, measuring a chords length. Multiple digital images can be used as initial data. Pre-processing is used to remove noise, sharpen and improve contrast using Adaptive Contrast-Limiting Histogram Equalization (CLAHE). The next step is grain segmentation. A combination of distance transform and adaptive watershed binarization is used. Binary images filtration based on the operations of mathematical morphology is provided. Contour analysis is used to determine grain boundaries. The study’s results of the entire rating scales and on the real metallographic images are presented. High efficiency of an algorithmic support is confirmed by the experiments. The software implementation has the following main features: the ability to calibrate the actual grain size, automatic or manual image preprocessing, grain size analysis with saving the results as a report in jpg format. Batch processing provides the ability to download images for processing with the same type of algorithm.

Strengthening Thin-Walled Knives with Nitrogen Plasma

Introduction. The completed developments are aimed at creating a new technology for increasing the wear resistance of a thin-walled instrument of complex configuration made of steel 65G for cutting beets at sugar enterprises. The most important requirement to improve the operability and durability of such a tool is the preservation of its profile and cutting edge during operation. Materials and Methods. A new developed equipment and technological process of strengthening using low-temperature nitrogen plasma were used to solve this problem. There have been determined optimal processing parameters that ensure the formation of a quasi-morphic structure on the friction surface that reduces the defectiveness of the cutting edge after its machining and also provides a process of self-sharpening due to tool strengthening on one side. Results. The comparative studies of the friction surface of products after operational tests have shown that their resistance increase significantly when strengthening both new and used products. This is determined by the nature of the quasi-morphic structure formed and the specific relief in friction on the working surface. Discussion and Conclusion. To describe the new process of strengthening thin-walled products, the structure formation on the friction surface was analyzed in detail with the use of metallographic images and its phase relationship variability was estimated by the optic-mathematical analysis of various zones (compression and vacuum) formed as friction bands. This was done trough modeling with the estimation of the distribution density of the conditional colors of the analyzed fragments.

ESTIMATION OF LOCAL INHOMOGENEITY OF MULTILAYER NANOSTRUCTURED ZrN/ZrO2 COATING

Multilayer nanostructured ZrN/ZrO2 coatings were applied to increase the operational resistance of various machinery parts by using vacuum-arc deposition in Bulat-type facility. To describe the structure formation, a new approach based on optical-mathematical method for processing metallographic images is proposed. The structure formation of the multilayer coating with an assessment of the degree of its inhomogeneity and diffusion processes between the layers is studied. For a reliable assessment, the changes in the horizontal and vertical directions of the images with the choice of optimal intervals were comparatively analyzed. It has been found that the most stable results are achieved using 20 and 25 dots (pixels).

An Improved Adaptive Weighted Mean Filtering Approach for Metallographic Image Processing

Background As noise brings great error in the analysis of metallographic images, an adaptive weighted mean filtering method proposed to overcome the shortcomings of the standard mean filtering method. Methods The method used to detect the pulse noise points in the image, and then the modified mean method used to filter out the detected noise points. Patents on metallographic image processing have discussed for the development of the proposed methodology. Results It is shown that filter window can be filtered in comparison with the conventional 3×3, 5×5 and 7×7 filt window to reduce noise detection and reduce the complexity of the weight calculation. Conclusion It can be concluded that this method can better protect the details of the image, has better filtering effect than the standard mean filtering, and its processing speed is faster than the median filtering of the large window, which has profound significance for the edge detection and processing of the metallographic image.