Characterisation of high temperature materials PDF

INTRODUCTION

Within the last ten years, the information technology revolution has come about as a result of, and also has been fuelled by, the increased accessibility of computing power. Most scientific instruments, industrial equipment and domestic appliances now rely on computers to operate, and new computers are cheaper, readily available and user-friendly. Thus in scientific and engineering applications, quantification is now used routinely whereas, before it might have been time-consuming and expensive. There are many scientific techniques that are benefitting from these developments including image analysis, the subject of this chapter and typical of one of the many quantitative techniques now used in materials' science. While traditionally used in materials' science to classify microstructures and so to help establish microstructure-property relationships, image analysis also has widespread usage in such diverse fields as data collection/processing for driverless vehicles, enhancement of images for the media, and the recognition of shapes and sizes in automated quality control and security systems.

An objective of image analysis in materials' science is to be able to measure parameters from one or more fields of view obtained from sources such as photographs and direct microscope images. This can readily be achieved either by using a video camera to capture an image or by processing the imaging signal directly from a microscope, and in either case then feeding the resulting signal via an interface to an image analyser. The image is thus digitised into small discrete areas (picture elements or pixels), each of which has a measured brightness (intensity or, in mono-chromatic systems, shades of grey). To a first approximation, the greater the number of pixels and available intensity levels, the higher the resolution of the digitised image and the better the potential discrimation of features within the image. Following digitisation, the image can be enhanced, if required, and then analysed. Note that a distinction should be made at the onset between image analysis (the measurement of features within a field of view) and image enhancement (manipulation of the original input digitised image in order to improve the appearance of the image and/or to facilitate the detection of features). The data remain unaffected in the former whereas the data (values of intensity for pixels and/or the number of pixels defining a feature) may be altered considerably in the latter, hence possibly affecting what subsequently is measured.

Having identified and measured features of interest, the final stages are statistical interpretation and stereology (mathematical manipulations used to transform the twodimensional data acquired back into threedimensional representations characteristic of the original microstructure) .

This chapter considers the various stages of image analysis. Further details can be found in the references at the end of the chapter.

Edited by: P. Spilling