Phenomenal advances in optical materials and signal processing hardware have led to smaller, feature-packed digital cameras capable of taking an almost perfect picture. While very little photographic knowledge is expected of the end user, designers and integrators of these 'digital platforms' require a thorough understanding of all aspects of 'imaging technologies.' With increasing demands for precision, speed, and data throughput, application developers need to be familiar with all related concepts, from capturing photogenerated electrons through interfacing front-end electronics to processing image data.
In machine vision, the 'vision' part of the work is done by the image sensor and algorithmic work by a 'machine.' Application developers and systems integrators must make decisions on how the target scene is to be captured (acquisition) while preserving the required attributes (features). Once the characteristics of the image acquired are assessed, manipulation operations (preprocessing) are to be chosen to facilitate the extraction of a predefined set of features from the acquired image (image analysis). Tasks in these three strands of machine vision require hardware knowledge of the 'front-end electronics' as well as analytical concepts of processing image data. The following chapters develop a framework for acquisition and preprocessing on the premise that these two strands are generic.
Apart from a small set of basic measurement operations, analytical and algorithmic work for image analysis, in contrast, is application specific, in that the specific features to be extracted dictate follow-up treatment of the preprocessed image data. The expanding literature on image analysis covers a variety of feature extraction and measurement techniques. An outline description of the first two strands and their linkage to the third is given here to underline their functional relation and to set a framework for the following chapters in the context of the well-documented literature on the general background of digital image processing.