Automated measurement of sintering degree in optical microscopy through image analysis of particle joins

Abstract

In general terms, sintering describes the bonding of particles into a more coherent structure, where joins form between packed particles, usually as a result of heating. Characterization of sintering is an important topic in the fields of metallurgy, steel, iron ore pellets, ceramics, and snow for understanding material properties and material strength. Characterization using image analysis has been applied in a number of these fields but is either semi-automatic, requiring human interaction in the analysis, or based on statistical sampling and stereology to characterize the sample. This paper presents a novel fully automatic image analysis algorithm to analyze and determine the degree of sintering based on analysis of the particle joins and structure. Quantitative image analysis of the sintering degree is demonstrated for samples of iron ore pellets but could be readily applied to other packed particle materials. Microscope images of polished cross-sections of iron ore pellets have been imaged in their entirety and automated analysis of hundreds of images has been performed. Joins between particles have been identified based on morphological image processing and features have been calculated based on the geometric properties and curvature of these joins. The features have been analyzed and determined to hold discriminative power based on a statistical evaluation using the Welch t-test and display properties consistent with sintering theory and results from traditional pellet diameter measurements on the heated samples.

Automated analysis of platelet microstructures using a feature length orientation space

“…Segmentation approaches are also extremely popular and allow properties of individual features to be measured [19,20]. In recent years, machine learning methods, particularly those based on convolutional neural networks and deep learning have become popular for segmentation and other inspection problems [21][22][23][24][25][26].…”

A new adaptive switching median filter for impulse noise reduction with pre-detection based on evidential reasoning

“…It means that a pixel with an extreme value should not be absolutely recognized as the noise because it might be the signal actually. Furthermore, Dempster's rule of combination in (5) has the problem of one ballot veto when one BBA is assigned 1 on one singleton θ i ( θ i ∈ ), while 0 on other singletons. That is, if m 1 (N) = 1 , i.e., m 1 (S) = 0 , no matter what m 2 is, the combined BBA has m (S) = 0 , which indicates the center pixel cannot be the signal.…”

“…The intensity of a pixel corrupted by the impulse noise tends to be much higher or lower than those of its uncorrupted neighbors. The impulse noise dramatically influences the image quality and makes images unsuitable for subsequent human understanding or image processing such as the edge detection [2] , segmentation [3] , object recognition [4] , image analysis [5] and image understanding [6] .…”

Discrimination of pores and cracks in iron ore pellets using deep learning neural networks

“…In another study, it was possible to perform a quantitative analysis of the degree of sintering from images obtained. (Nellros et al, 2015) However this technique, besides being destructive, provides only 2D information, limiting the study and the visualization of the internal and complete structure of the pellet. For this reason, x-ray microtomography (microCT) has been used with the advantage of being a non-destructive technique and having the ability to generate three-dimensional images (3D).…”