No abstract

The study of the relationship among the manufacturing process, the structure and the property of materials can help to develop the new materials. The material images contain the microstructures of materials, therefore, the quantitative analysis for the material images is the important means to study the characteristics of material structures. Generally, the quantitative analysis for the material microstructures is based on the exact segmentation of the materials images. However, most material microstructures are shown with various shapes and complex textures in images, and they seriously hinder the exact segmentation of the component elements. In this research, machine learning method and complex networks method are adopted to the challenge of automatic material image segmentation. Two segmentation tasks are completed: on the one hand, the images of the titanium alloy are segmented based on the pixel-level classification through feature extraction and machine learning algorithm; on the other hand, the ceramic images are segmented with the complex

No abstract

Microstructure of a material stores the genesis of the material and shows various properties of the material. To efficiently analyse the microstructure of a material, the segmentation of different phases or constituents is an important step. However, in general, due to the microstructure’s complexity, most of segmentation is manually done by human experts. It is challenging to automatically segment the material phases and the microstructure. In this work, we propose a method which combines the the dilation operator, GLCM (gray-level co-occurrence matrix), Hough transform and DBSCAN (density-based spatial clustering of applications with noise) for phases segmentation in the examples of certain material of eutectic HfB2-B4C ceramics. In the segmented regions, the further analysis for the microstructural elements is done with DBSCAN. The experimental results show that the proposed method achieves 95.75% segmentation accuracy for segmenting phases and 86.64% correct classification rate for the microstructure in the segmented phases. These experimental results show that our method is effective for the difficult task of the both segmentation and classification of the microstructural characteristics.

No abstract