Previous: dir Up: ../plot79.html Next: graph

Table of contents

The following is a glossary of common phrases used in computer graphics. Many have been taken, sometimes with minor modifications or additions, from the 1979 CORE report and the 1984 GKS draft proposal.


          The elimination of visible stair-step effects on  lines  drawn
          on  a  raster display by distributing the intensity error into
          neighboring pixels which are not directly intersected  by  the


          A plane parallel to the view plane and displaced from it along
          the  view  plane normal by the back distance which is measured
          from the view reference point.  When back clipping is enabled,
          portions  of  the  object  which lie behind this plane are not


          A  display device configuration in which the intensity of each
          pixel is described by a single bit  (usually)  in  a  computer
          memory.  Typical examples are low-cost raster displays and dot
          matrix printers.


          A  world  coordinate  object  is projected onto the view plane
          with either a perspective  or  a  parallel  projection.    The
          projection  of  an  object  onto  the  view  plane is found by
          passing lines through each point of  the  object  and  finding
          their  intersection  with  the  view plane.  For a perspective
          projection, all lines emanate from a common  point,  which  is
          the  center  of  projection.    For a parallel projection, the
          lines are all parallel to a specified vector,  in  which  case
          the  center  of  projection may be said to be at infinity.  In
          the CORE system, the  center  of  projection  for  a  parallel
          projection  is  simply  a point such that all projection lines
          are parallel to the line from it to the view reference  point.
          If the projection lines are parallel to the view plane normal,
          the parallel projection is called orthographic, or  sometimes,
          axonometric.    This includes the isometric projection used in
          drafting.  Skewed parallel projections are obtained  when  the
          projection  lines  are  not parallel to the view plane normal.
          Examples    include    the    Cabinet  and  cavalier  drafting


          Removing parts of display elements that lie  outside  a  given
          boundary, usually a window or viewport.


          An integer index into the color lookup table.


          A  display device dependent table in which the entries specify
          the values of red, green,  and  blue  intensities  defining  a
          particular color.  This capability is available on most better
          quality  color  raster  displays,  and  serves  two  purposes.
          First,  only  a  limited number of memory bits (typically 2 to
          16, occasionally up to 24) need be allocated to each pixel  on
          the  display,  but these form an integer index into the lookup
          table which can specify red, green, and blue intensities  much
          more precisely, offering a wider range of possible colors than
          would   otherwise  be  available.    Second,  the  display  is
          dynamically updated from the lookup table which can be changed
          in  less  than  one  refresh  cycle  to  radically  alter  the
          appearance  of  the  displayed  image;  this  is of particular
          importance in image processing and enhancement.  If the memory
          values  specified the color mixture directly, this much larger
          amount of memory could not be updated nearly as quickly.


          The  description of color in terms of its constituent additive
          primary intensities of red, green, and blue (RGB) (as used  by
          television  monitors),  or  sometimes  in terms of subtractive
          primaries cyan, magenta, and yellow (CMY) (as  used  by  color
          printing  devices).    These  two  are  hardware oriented, and
          specification of the mix of RGB or CMY to obtain a  particular
          color  is  rather  non-intuitive.  Other models which are more
          natural for a human are the  hue,  lightness,  and  saturation
          (HLS)  model  and  the  hue, lightness and value (HLV) models.
          The CORE system allows color specification in terms of  either
          the RGB or HLS models.


          A device independent sequential file which  can  be  read  and
          written  by  the  CORE  system.   It may be used for long term
          storage    and    transmittal   and  transferral  of  graphics
          information.  In the PLOT79 implementation of Metafile output,
          there is a single device driver which can be used in place  of
          any  other  device  driver.    The  PLOT79 Metafile Translator
          program can be used with  any  device  driver  to  graphically
          display  the  contents  of a Metafile, and the Metafile Dumper
          program can be used to print summaries of the  contents  of  a


          A proposal for a 2-D and 3-D graphics standard developed by an
          international  working  group (the Graphics Standards Planning
          Committee)  from  the  early  1970's  with  the  final  report
          published  in 1979 (ACM SIGGRAPH Computer Graphics, Volume 13,
          Number 3, August 1979).  The  name  PLOT79  commemorates  this


          CORE system value that defines the current drawing position in
          world  coordinates.    It  is  set  to the origin of the world
          coordinate system at CORE system  initialization  time.    Its
          value  is  affected by calls made to the functions that create
          output primitives.


          Graphics  Kernel  System  -  a draft ANSI and ISO 2-D graphics
          standard    originally    developed    by    the  West  German
          Standardization  Institute,  DIN, in 1978.  It is published in
          ACM SIGGRAPH Computer Graphics, Volume  18,  Number  2,  March


          The device dependent part of a CORE implementation intended to
          support  a  graphics  device.    It  generates device specific
          output in response to  device-independent  requests  from  the
          CORE routines.


          A device (for example, refresh display, storage tube  display,
          pen  plotter,  film  recorder,  dot  matrix  printer) on which
          display images can be represented.


          In  a  display device, that medium on which display images may


          A  technique  for  simulation  of  intensity ranges on display
          devices    which    have    substantially   limited  intensity
          possibilities  (usually  just  black (zero) and white (full)).
          One such method is the halftone process used for image display
          in  books  and newspapers, in which small rectangular areas of
          paper are printed with areas  of  ink  in  proportion  to  the
          desired  intensity.    Techniques  for  bit-map and dot matrix
          displays which support only black/white choices at each  pixel
          attempt  to redistribute over neighboring pixels the intensity
          error made at each pixel by having to color it  totally  black
          or totally white.


          Enhancement of the interior of closed 2-D polygonal regions by
          filling them with solid color, patterns, or a hatch style.


          A text attribute determining the shape of characters and their
          relative positioning.


          A display technology in which the  intensities  and  color  of
          each  pixel  on  the display surface is maintained in computer
          memory and  dynamically  redisplayed  in  raster  line  order,
          usually  at  rates  from  30  to  100  times  per second, on a
          television  monitor.    By  comparison,  broadcast  television
          refresh  rates  are  25 to 30 times per second, and movie film
          rates are 24 frames per second.   With  an  appropriate  decay
          time  of  the  television phosphor and refresh rates somewhere
          above 40 frames per second, it is possible  to  obtain  motion
          which is flicker-free to the human eye.


          A plane parallel to the view plane and displaced from it along
          the  view plane normal by the front distance which is measured
          from the  view  reference  point.    When  front  clipping  is
          enabled,  portions  of  the  object which lie in front of this
          plane are not plotted.


          An algorithm for the display of mesh or wire frame surfaces in
          which lines on the mesh which lie behind features between them
          and  the  viewpoint  are  not shown.  These algorithms usually
          work by scanning the mesh line segments in order from front to
          back comparing each segment against upper and lower visibility
          horizons which are the silhouette of the object to that  point
          in the scan.  The advantages are that the scan time is usually
          linear in the number of points on the mesh, and  the  case  of
          elevation  data  on  a grid of points, for which the method is
          well suited, occurs very frequently, and is easily represented
          in  a  computer  program  by  a rectangular array of elevation
          values.  The disadvantages are that the method  is  only  with
          difficulty  extended  to multiple surfaces, contours on meshed
          surfaces, or multiple-valued surfaces, and that it deals  with
          individual  line segments, rather than surface patches, making
          smooth shading impractical.


          An algorithm for the display of solid objects in which surface
          portions are hidden by other parts of the object between  them
          and  the  viewer.    Surface portions which face away from the
          viewer   are  similarly  hidden.    The  objects  are  usually
          represented  by  planar  polygonal  surface patches (or curved
          patches in newer work) and information about how  the  patches
          are  connected  to form the surface.  The advantage of this is
          that the resulting display can be  much  more  life-like  than
          wire  mesh  drawings, particularly if the visible parts of the
          surface are given different  intensities  according  to  their
          shading  by  a  light  source.    Surface  texturing  is  also
          possible.  The disadvantage is that  sophisticated  algorithms
          are  needed  to  avoid  the  N-squared computation growth from
          comparing each patch  with  every  other  one  to  find  which
          portion  is  visible,  and  sophisticated  data structures are
          necessary to  represent  the  surface  patch  orientation  and
          connectivity information.  Another disadvantage is the general
          inability to deal with translucent  and  transparent  objects,
          for which the ray tracing algorithm is required.


          A quaint name for the front clipping plane.


          The  characteristic associated with a color names such as red,
          yellow, green, blue, etc.  Hue is a gradation of  color.    In
          the  CORE  HLS model, blue is at 0 and 1, red at 1/3, green at
          2/3, so that the hues form a unit cycle.


          A  branch  of  computer graphics involving the enhancement and
          analysis of images obtained from raster scanning devices  such
          as  television  cameras,  computer-assisted  tomography  (CAT)
          scanners, satellites  such  as  LANDSAT,  spacecraft  such  as
          Mariner and Voyager, and so on.


          The characteristic that allows color to be ranked on  a  scale
          from light to dark.


          A directed line segment.  In the CORE system, it is a  visible
          line segment from the current position to the specified point,
          after which that point becomes the new current point.


          An  attribute of lines in the CORE system; lines may be solid,
          or dashed and dotted in a variety of combinations.


          A symbol with a specified appearance which is used to identify
          a particular location.


          A  coordinate  specified  in  a  device independent coordinate
          system normalized to the range 0 to 1.


          The  2-D  or 3-D Cartesian space which represents the viewable
          region of an arbitrary display device.  In the case of 3-D NDC
          space,  the  coordinate system is left-handed, with the origin
          at the lower left front face of the unit cube, x  positive  to
          the right, y positive upward, and z positive into the cube.


          A less-common name for pixel.


          The  smallest  element  of  a  display  surface  that  can  be
          independently assigned a color or intensity.  The name  is  an
          acronym  for  "picture element".  It is conventionally applied
          primarily to raster scan displays.


          A  closed  planar  figure  bounded  by straight line segments.
          Although the CORE system permits this to be specified  in  3-D
          coordinates, it must still lie in a single plane.


          A set of connected line segments.


          A  set  of locations, each to be indicated by the same type of
          marker symbol.


          A  picture  element,  such  as a line, marker, or text string,
          having a specific appearance.  The CORE system  supports  five
          output  primitives:    line, polyline, marker, polymarker, and
          text; the raster extensions to the CORE system add  a  polygon
          primitive.    The  GKS  system  supports polyline, polymarker,
          text,    fill   area,  cell  array,  and  generalized  drawing


          Computer graphics in which a display image is composed  of  an
          array of pixels arranged in rows and columns.


          A sophisticated (and computationally expensive) algorithm  for
          constructing  high-quality  hidden  surface  images  including
          reflection, refraction, and shadowing.  For each pixel in  the
          image,  the  algorithm  traces a light ray into the scene from
          the pixel to the object and after branching due to  reflection
          and refraction, back to the original light source(s).


          The graphical display of objects rapidly enough that the scene
          can  be  changed  to give the appearance of smooth motion to a
          human viewer in front of the display.   Until  recently,  such
          displays  have  been  very  expensive  (in  the range of fifty
          thousand    to  several  million  dollars).    Television  and
          commercial  film computer graphics motion sequences which have
          become increasingly common from film makers such as  Lucasfilm
          and Disney are actually not done in real time.  Each frame may
          represent many minutes  or  even  hours  of  computation,  and
          require  film  exposures  of  several seconds for each primary


          Turning all or part of a display image about an axis.


          The characteristic which describes the extent to which a color
          differs  from  a  gray of the same lightness.  In the CORE HLS
          model, zero saturation corresponds to maximum white content at
          that lightness level, and unity to full saturation.


          Enlarging or reducing all  or  part  of  a  display  image  by
          multiplying  the coordinates of display elements by a constant


          A  refresh  display  in which a command list (usually called a
          display list) of graphics primitives is repeatedly scanned and
          redrawn  directly on the face of the display tube.  This is in
          contrast to a frame buffer  display  in  which  the  image  is
          displayed by a raster scan.  The advantage of a stroke display
          is that  real-time  dynamic  graphical  display  is  possible,
          provided  that  the  display  list  is  not  too long, and the
          phosphor decay rate is short enough  to  avoid  ghosting,  and
          long  enough to avoid flicker.  The main disadvantages are the
          expense of the hardware  required  for  adequate  performance,
          particularly  if  color  is required, and with few exceptions,
          the inability to handle shading and polygon  filling  in  real


          An output primitive consisting of a character string.


          The  application of a constant displacement to the position of
          all or part of a display image.


          A  signed  quantity  which  is  the  distance  from  the  view
          reference point to the view plane measured in the direction of
          the view plane normal.


          A 3-D vector specified relative to the view  reference  point.
          The  view  plane containing the window is perpendicular to the
          view plane normal at a distance from the view reference  point
          specified  by  the  view  plane  distance.   The origin of the
          window coordinate system  is  at  the  point  where  the  line
          through  the  view  reference point parallel to the view plane
          normal pierces the view plane.  In the CORE default case,  the
          view  distance is zero, and the view reference point then lies
          on the window at the coordinate system origin.


          A convenient known point on or near the object being viewed.


          A 3-D vector whose projection onto the view plane defines  the
          up direction in the window.


          An application program specified  part  of  normalized  device
          coordinate  space.  This is restricted to a rectangular region
          in 2-D NDC  space,  or  a  truncated  rectangular  prism  with
          parallel front and back faces in 3-D NDC space.


          An application program specified rectangular  portion  of  the
          view  plane  in which the view of the object appears according
          to the viewing parameters and projection type selected.  It is
          mapped  onto  the  viewport in NDC space.  If the NDC space is
          3-D, the front clipping plane is mapped onto the  front  plane
          of  the 3-D NDC viewport, and the back clipping plane onto the
          back plane of the viewport.  If the aspect ratio of the window
          is  different than that of the viewport, then the view will be
          stretched or shrunk accordingly.


          A  device  independent 3-D Cartesian coordinate system used by
          the application program for  specifying  graphical  input  and
          output.    The  units are arbitrary, for example, light years,
          meters, furlongs, time, inches, parsecs, miles, weight, and so
          on.    In the CORE system, the coordinate system may be chosen
          either right-handed or left-handed.


          A quaint name for the back clipping plane.


          A refresh buffer in which,  for  each  pixel  on  the  display
          surface,  in addition to color values, is maintained the depth
          (z coordinate) of the object mapping  into  that  pixel.    By
          retaining  only the smallest z value, which corresponds to the
          object closest  to  the  window,  a  hidden  surface  view  is


          Enlarging or reducing all  or  part  of  a  display  image  by
          multiplying  the coordinates of display elements by a constant
          value.  When this feature is implemented on a  raster  device,
          it  normally  allows  only integral enlarging factors of a 2-D
          display, most frequently in the range 1 to 16.  Higher quality
          raster displays may permit different scale factors for the two