Glossary

This glossary contains descriptions of significant terms that appear in this book.

affine transformation: Geometric image transformation, such as translation, scaling, or rotation.
band: The set of all samples of one type in an image, such as all red samples or all green samples.
box filter: A low-pass spatial filter composed of uniformly-weighted convolution coefficients.
bicubic interpolation: Two-dimensional cubic interpolation of pixel values based on the 16 pixels in a 4 x 4 pixel neighborhood. See also bilinear interpolation, nearest-neighbor interpolation.
bilinear interpolation: Two-dimensional linear interpolation of pixel values based on the four pixels in a 2 x 2 pixel neighborhood. See also bicubic interpolation, nearest-neighbor interpolation.
binary image: An image that consists of only two brightness levels: black and white.
chain code: A pixel-by-pixel direction code that defines boundaries of objects within an image. The chain code records an object boundary as a series of direction codes.
cobble: To assemble multiple tile regions into a single contiguous region.
color space conversion: The conversion of a color using one space to another color space, such as RGB to CMYK.
components: Values of samples independent of color interpretation.
compression ratio: In image compression, the ratio of an uncompressed image data file size to its compressed counterpart.
data element: Primitive types used as units of storage of image data. Data elements are individual members of a DataBuffer array.
directed graph (digraph): A graph with one-way edges. See also directed acyclic graph (DAG).
directed acyclic graph (DAG): A directed graph containing no cycles. This means that if there is a route from node A to node B then there is no way back.
first-order interpolation: See bilinear interpolation.
high-pass filter: A spatial filter that accentuates an image's high-frequency detail or attenuates the low-frequency detail. Contrast with low-pass filter, median filter.
histogram: A measure of the amplitude distribution of pixels within an image.
Lempel-Ziv-Welch (LZW) compression: A lossless image coding method that scans the image for repeating patterns of blocks of pixels and codes the patterns into a code list.
low-pass filter: A spatial filter that attenuates an image's high-frequency detail or accentuates the low-frequency detail. Contrast with high-pass filter, median filter.
median filter: A non-linear spatial filter used to remove noise spikes from an image.
nearest-neighbor interpolation: Two-dimensional interpolation of pixel values in which the amplitude of the interpolated sample is the amplitude of its nearest neighbor. See also bicubic interpolation, bilinear interpolation.
perspective warp: An image distortion in which objects appear trapezoidal due to foreshortening.
projective warp: See perspective warp.
quantization: The conversion of discrete image samples to digital quantities.
ROI: Abbreviation for region of interest. An area or pixel group within an image that has been selected for processing.
run-length coding: A type of lossless image data compression that scans for sequences of pixels with the same brightness level and codes them into a reduced description.
Sobel edge detection: A spatial edge detection filter that detects edges by finding the gradient of an image.
square pixel: A pixel with equal height and width.
thresholding: A point operation that maps all the pixel values of an image that fall within a given range to one of a set of per-band constants.
transform coding: A form of lossy block coding that transforms blocks of an image from the spatial domain to the frequency domain.
trapping: An image manipulation technique used in printing that uses dilation and erosion to compensation for misregistration of colors.
unsharp masking: An image enhancement technique using a high-frequency accentuating filter.
zero-order interpolation: See nearest-neighbor interpolation.