Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
45 Cards in this Set
- Front
- Back
Image Generation |
1. first know where images/sources comes from 2. the images converted to digital format: digital compositing |
|
Sources of elements 1. hand painted/ human generated elements |
examples: - simple black & white matte - photo-realistic matte painting of a real/imaginary setting Past: artwork was generated using traditional painting methods without using computer Today: hand-paint elements directly in the computer using paint programs |
|
Sources of elements 2. computer generated images |
elements that have been created/rendered using a specialized 2D/3D computer animation package may contain scanned/painted elements |
|
Sources of elements 3. images that have been digitized into the computer from some other source |
eg: film / video |
|
pixels |
digital images are stored as an array of individual dots |
|
channels |
a color image as a layered collection of simpler images |
|
digital image |
a rectangular array of pixel each pixel has a characteristic color associated with it : RGB components |
|
components |
one of the elements that is used to define the color of a pixel eg: red |
|
RBG components/model/color space |
by using a combination of these 3 primary colors at different intensities, we can represent the full range of the visible spectrum for each pixel
* 3 layer combination of primary colored channels |
|
sample color image |
the red, green, blue channel |
|
the three channels |
transparent slides that could also be considered as monochrome (black &white) images on their own eg: red channel in grayscale |
|
motivation for RGB Channel |
1. single pixels is too small to deal with individually 2. channels gives additional control 3. allows for the use of techniques based on optical compositing 4. the 3 channels can be manipulated/handled separately |
|
why use RGB channels? |
manipulating and combining the individual channels can be used to do other operations like color correction and matte extraction |
|
resolution |
the amount of data that is used to capture an image width * height a larger number of pixels in an image allows for finer image detail |
|
types of resolution |
1. spatial 2. color 3. remporal |
|
spatial resolution |
a measure of the amount of data used to capture an image: pixel count |
|
pixel count |
primary measurement for an image's resolution |
|
color/dynamic/chromatic resolution |
the amount of data allocated for specifying the value of an individual color of an image |
|
bit depth |
number of bits per pixel a way of measuring the color resolution of a given image a large number of bits allows for finer variations in color |
|
bit depth for image store
|
8 bits per channel : 24-bit image each component of a pixel can have 256 intensities 3 components together can represent about 16 million colors |
|
8 bits per channel |
after reducedto 4 bits per channel |
|
quantizing |
noticeable delineations between various bands of colors when we do not have the ability to specify enough unique color values fro smooth transitions between different shading also referred as: banding, contouring, posterization eg: 3 bits per channel: only 8 shades |
|
Normalized Values |
take different numerical ranges and normalized to floating-point non-integer in the range of 0-1 eg: 8 bits per channel-->range 0-255 divide each value by 255 to come up with the normalized value - (255, 100, 0)-->(1.0, 0.39, 0) |
|
benefits of normalized values |
- do not have to worry about bit depths - makes math operations between images easier - (1, 1, 1) white pixel - (0, 0, 0) black - (0.5, 0.5, 0.5) 50% gray |
|
floating point image representation |
rendered into a floating point image representation --> true values pixel values above 1 are not able to represented wll |
|
Floating point image with values reduced by half |
fixed point images with values reduced bu half |
|
working with float mode |
- not yet overall standard for digital compositing work - large file size like 32 bits per channel may not work well - less data loss if the image undergoes multiple manipulations |
|
high dynamic range imagery(HDRI) |
imagery captured from the real world that contain a large range of brightness values |
|
HSV Color Representation/model/color space - HSB (brightness) - HSL (lightness) |
hue, saturation, value of a pixel - more intuitive than RGB model - alternate method of representing data - generally not used to store images - used commonly to manipulate colors |
|
hue of a pixel = its basic color image: individual hue component/channels |
range of 0 -360 |
|
saturation of a pixel = the brilliance/purity of the specific hue that is present in the pixel image: individual saturation component/channels --> most of the colors in this image are close to white/saturated |
color are fully saturated at the edge saturation decreases as you move to the center of the wheel |
|
value of a pixel = brightness of the color image: individual value component/channels-->the beak of the parrot is brighter than background |
value is along the vertical axis lowest value - black highest value - white |
|
YUV Color Representation/model/color space |
Y = luminance (intensity) 亮度 U & V : chrominance values 色度 - not intuitive - combined hue & saturation common in video |
|
YUV component image |
|
|
additional image channel : alpha / transparency / matte |
the portion of a four-channel image that is used to store transparency of various pixels |
|
compression - run length encoding - lossy compression - chroma subsampling |
high resolution images can take up huge amounts of space --> need to compress 1. lossless compression 2. lossy compression |
|
lossless compression |
a method of compressing and storing a digital image in such a fashion that the original image can be completely reconstructed without any data loss |
|
lossy compression |
a method of compressing and storing a digital image in such a fashion that it is impossible to perfectly reconstruct the original image |
|
run-length encoding (RLE) |
compress image without loss of information 12:0 12:0 5:0, 2:1, 5:0 reduced the amount of information by about ..% (new characters/original) |
|
Lossy compression |
information loss when image is compressed JPEG compression advantages: - maintains brightness and contrast information (human eye is very sensitive to these components), removing some color information - the user can specify the amount of compression that is used when storing the file * not good for synthetic image |
|
chroma subsampling |
- used in video, where video is generally kept in YUV color model the eye is more sensitive to changes in luminance than in chrominance one compression method would be to reduce the amount of chrominance store/transmit 2 chrominance samples for every 4 luminance samples 4:2:2 4:1:1 4:4:4 (full quality signal) 4:0:0 (no color information, B&W image) |
|
4:1:1 compression |
encoding the data needed to represent an image by sampling Y for every pixel but removing every other UV pixel in both horizontal and vertical directions 1 Y/chrominance sample for every 4 UV/luminance samples eg: in 16 frames 4:1:1 --> 4Y 1ca 1cb --> 16Y 4ca 4ba |
|
4:2:2 compression |
encoding the data needed to represent an image by sampling Y for every pixel but removing every other UV pixel in horizontal direction |
|
4:4:4 compression |
encoding the data needed to represent an image by sampling Y and UV for every pixel in the image (full quality signal) |
|
image sequence compression |
- store sequence of images - instead of storing every single frame in the sequence ---> store master frame, then only record areas of different from the master frame - significant reduction in data size |