A typical pattern-recognition task in the image-processing pipeline includes the following steps: signal acquisition, data preprocessing, pixel registration, image fusion, region segmentation, feature extraction, pattern classification, and decision making. Each step is briefly described in this chapter except for image fusion, which will be addressed in Chapters 6 to 10.

5.1 Preprocessing

5.1.1 Image acquisition

Image acquisition is the first task in the processing pipeline; it refers to the collection of signal information to develop the image. Images may be collected upon a research plan or preanalysis of results from an imaging sensor. Given a series of images, it is important to store the images in a database that is diverse and representative of OCs. For face-recognition research, the following aspects may be considered during the image-acquisition stage. The information can be collected and recorded during the image-acquisition stage, and then stored as metadata for ease of indexing, access, search, and retrieval.