CCS 2208 COMPUTER GRAPHICS

BY

JULIET MOSO
 What is Computer Graphics?
• Creation, Manipulation, and Storage of geometric objects
(modeling) and their images (rendering)

• Typical graphics system comprises of a host computer

with support of fast processor, large memory, frame buffer
and
• Display devices (colour monitors),
• Input devices (mouse, keyboard, joystick, touch screen,
trackball)
• Output devices (LCD panels, laser printers, colour
printers. Plotters etc.)
• Interfacing devices such as, video I/O, TV interface etc.
 24/01/2025 Lecture 1 3

Graphics Definitions
• Point:
• a location in space, 2D or 3D sometimes denotes one pixel
• Line
• straight path connecting two points
• Vertex
• point in 3D
• Edge
• line in 3D connecting two vertices
• Polygon/Face/Facet
• arbitrary shape formed by connected vertices, fundamental unit of
3D computer graphics
• Mesh
• set of connected polygons forming a surface (or object)
24/01/2025 Lecture 1 4

Graphics Definitions
• Rendering : is the process of generating an image from
a 2D or 3D model (or models in what collectively could be
called a scene file), by means of computer programs.
Also, the results of such a model can be called a
rendering.

• Framebuffer : a large block of memory that stores the

color values for all the pixels on the screen.
 What drives computer graphics?
• Movie Industry
• Leaders in quality and artistry
• Big budgets and tight schedules
• Defines our expectations

• Game Industry
• The newest driving force in CG due to volume
and Profit
• Focus on interactivity
• Cost effective solutions
 What drives computer graphics?
• Medical Imaging and Scientific Visualization
• Tools for teaching and diagnosis
• No cheating or tricks allowed
• New data representations and modalities
• Drive issues of precision and correctness
• Focus on presentation and interpretation of data
• Construction of models from acquired data

Nanomanipulator, UNC Joe Kniss, Utah Gordon Kindelman, Utah

 What drives computer graphics?
• Computer Aided Design
• Drives the high end of the hardware market
• Integration of computing and display resources
• Reduced design cycles results in faster systems in a shorter
timeframe

• Education and Training

• Models of physical, financial, social systems
• Comprehension of complex systems

• Computer Art
• Fine and commercial art
• Performance Art
• Aesthetic Computing
 Historical Milestones
• 1960’s:
• Early theoretical development, mainly limited to research and military
• 1962: Sketchpad (Ivan Sutherland)
• 1970’s:
• ‘Traditional’ graphics pipeline developed
• Driven by money from military simulation and automotive design industries
• 1980’s:
• Many important core algorithms developed
• Visual quality improved driven by demands from entertainment (movie) industry
• 1985: Rendering Equation (James Kajiya)
• 1990’s:
• Advanced algorithms developed as graphics theory matured
• Broader focus on animation, data acquisition, NPR, physics…
• 1995: Photon Mapping (Henrik Jensen)
• 2000’s:
• Photoreal rendering evolves to the point of being able to render convincing
images of arbitrarily complex scenes on consumer hardware
• Merging of computer graphics and computer vision
• Cheap graphics hardware with vast capabilities, driven largely by video game
industry
 Image Processing
• Some computer graphics operations involve
manipulating 2D images (bitmaps)

• Image processing applies directly to the pixel grid

and includes operations such as color correction,
scaling, blurring, sharpening, etc.

• Common example include digital photo

processing and digital ‘painting’ programs (Adobe
Photoshop…)
Image Synthesis
• Image synthesis or image generation refers more
to the construction of images from scratch, rather
than processing of existing images

• Synthesis of a 2D image from a 3D scene

description is more commonly called rendering
 Photoreal Rendering
• Photoreal rendering refers to rendering a 3D scene in a
realistic way

• Modern photoreal rendering algorithms are essentially a

physically based simulation of light propagation and
scattering throughout a 3D environment

• This means that there is a ‘correct’ image that should be

generated, given an input data set. This allows the
subject of photoreal rendering to have a strong theoretical
basis (namely, the science of optics)

• Most modern photoreal rendering algorithms are based

on the classic ray tracing algorithm, that traces the path of
individual light rays starting from the eye and working
backwards to the light sources.
Non-Photoreal Rendering
• Non-photoreal rendering (NPR) refers to
rendering images in other ways…

• Sometimes, this is done to achieve aesthetic

goals such as artificial water colors, pencil
sketches, paint brushstrokes…

• Other times, the goal is to maximize the

communication of visual information, as in
scientific and medical visualization
NPR
Computer Vision
• Computer vision is sometimes considered as a
separate discipline from computer graphics,
although they share many things in common

• A central goal in computer vision is to take a set of

2D images (usually from a video or set of photos)
and infer from that a 3D description of what is
being viewed

• This is a very different process than rendering,

and is more of a form of artificial intelligence
 Animation
• An animation is just a sequence of individual images

• Basically, the subject of computer animation focuses on

how things change over time. Usually, this refers to
motion, but can also refer to other properties changing
over time.

• Physical simulation is a very powerful tool in computer

animation and can be used to generate believable
animations of rigid objects, deformable objects, gasses,
liquids, fracture, particle effects, and even explosions
and fire

• Computer animation also includes a large number of

techniques specifically developed to manipulate virtual
characters
Modeling
• Modeling refers to the techniques involved with creating,
scanning, editing, and manipulating 3D geometric data
• It is done by a human user with an interactive editing
program
• More complex objects, such as trees, can be constructed
with automatic procedural modeling algorithms
• 3D models are also often acquired from real world objects
using laser scanning or computer vision techniques
• Modeling also includes the use of curved surfaces and
other higher order primitives, which are often converted into
triangles using various tessellation algorithms
• It also includes mesh reconstruction for surface
simplification
• Modeling makes heavy use of computational geometry
 Display Technology
• Vector display (cathode ray tube)

• Raster displays
• CRT (cathode ray tube)
• LCD (liquid crystal display)
• TFT (thin film transistor)
• OLED (organic light emitting diode)
• Light valve
• Plasma
• HDR (high dynamic range: TFT / white LED hybrid)

• Film
• Print
 Raster Graphics
• Modern graphics displays are raster based
• They display a grid of pixels, where each pixel color can
be set independently
• Individual pixels are usually formed from smaller red,
green, and blue subpixels. If you look very closely at a
TV screen or computer monitor, you will notice the
pattern of subpixels
• Older style vector displays didn’t display a grid of pixels,
but instead drew lines directly with an electron beam
• Raster graphics are also sometimes called bitmapped
graphics
 Interlacing
• Older video formats (NTSC, PAL) and some HD formats
(1080i) use a technique called interlacing
• With this technique, the image is actually displayed twice,
once showing the odd scanlines, and once showing the
even scanlines (slightly offset)
• This is a trick for achieving higher vertical resolution at the
expense of frame rate (cuts effective frame rate in half)
• The two different displayed images are called fields
• NTSC video, for example, is 720 x 480 at 30 frames per
second, but is really 720 x 240 at 60 fields per second
• Interlacing is an important issue to consider when working
with video, especially in animation as in TV effects and
video games
• Computer monitors are generally not interlaced
 Framebuffer
• The framebuffer refers to the memory dedicated to storing
the image
• It would generally be a 2D array of pixels, where each
pixel stores a color (Note: pixel = picture element)
• Color is typically stored as a 24 bit RGB value. This offers
8 bits (256 levels) for red, green, and blue, for a total of
16,777,216 different colors
• Very often, additional data is stored per pixel such as
depth (z), or other info
• A framebuffer can just be a block of main memory, but
many graphics systems have dedicated framebuffer
memory with a direct connection to video scan-out
hardware and other special features
Primitives
• Complex scenes are usually built up from simpler objects
• Objects are built from individual primitives
• The most common and general purpose 3D primitive is
the triangle
• Points and lines are also useful primitives

• Four basic output primitives (or elements) for drawing

pictures:
• Polyline
• Filled POLYGONS (Regions)
• Ellipse (Arc)
• Text
• Raster IMAGE
Traditional Graphics Pipeline
• In the traditional graphics pipeline, each primitive is
processed through the following steps:
• Transformation
• Lighting
• Clipping
• Scan conversion
• Pixel processing
Transformation
• The transformation process refers to the linear
transformation from 3D space to a 2D viewing
space

• Each vertex position must be transformed from

it’s defining object space to the device
coordinates (pixel space)

• This often involves a combination of rotations,

translations, scales, and perspective
transformations
Lighting
• Lighting operations are applied to each vertex to
compute its color

• In more advanced rendering, lighting operations

are computed per pixel, rather than per vertex

• A variety of light types can be defined such as

point lights, directional lights, spot lights, etc.

• More advanced lighting operations can account for

shadows, reflections, translucency, and a wide
variety of optical effects
Clipping
• Some triangles will be completely visible on the
screen, while others may be completely out of view
• Some may intersect the side of the screen and
require special handling
• The camera’s viewable space forms a volume called
the view volume. Triangles that intersect the boundary
of the view volume must be clipped.
• The related process of culling refers to the
determination of which primitives are completely
invisible
• The output of the clipping/culling process is a set of
visible triangles that lie within the dimensions of the
display device
Scan Conversion
• The scan conversion (or rasterization) process takes
2D triangles as input and outputs the exact pixels
covered by the triangle

• Per-vertex data, such as color, is interpolated across

the triangle, so each pixel may have a unique color
 Pixel Processing
• The output of the scan conversion process is a bunch of
individual xy pixels, plus additional data per pixel such as
interpolated depth (z), color, or other information
• The pixel processing stage includes the operations that take
place per pixel to compute the final color that gets rendered
into the framebuffer
• Usually, the zbuffer technique is used to make sure that a
pixel is rendered only if it is not blocked by an existing
surface
• Other processing, such as texturing and transparency
operations happen per pixel
• In some systems, the entire lighting process is computed
per pixel, instead of per vertex
Scene Rendering
• With the traditional zbuffered graphics pipeline,
triangles can be rendered in any order without
affecting the final image
• Often, complex effects such as transparency, actually
do depend on the rendering order, and so may
require additional care
• Still, it makes a nice basic approach, and it’s the
approach taken by OpenGL and built into many
modern hardware graphics boards
• There are more advanced rendering algorithms (scan
line, ray tracing, etc.) that don’t render triangles one
at a time, and require the entire scene to be
processed.