Image Quantization, Halftoning, and Ditheringkucg.korea.ac.kr/education/2008_2/CNCE340/tutor_/08ImageProcessing.pdfGraphics Lab @ Korea University Image Resolution Intensity resolution

Graphics

Graphics Lab @ Korea Universitykucg.korea.ac.kr

Image Processing

고려대학교 컴퓨터 그래픽스 연구실

KUCG


Overview

Image RepresentationWhat is an image?

Sampling and ReconstructionKey steps in image processing

KUCG


What is an Image?

An image is a 2D rectilinear array of pixels

Continuous image Digital image

KUCG


What is an Image?



KUCG


What is an Image?



A pixel is a sample, not a little square!!

KUCG


Image Acquisition

Pixels are samples from continuous functionPhotoreceptors in eyeCCD cells in digital cameraRays in virtual camera

KUCG


Quantization

Artifact due to limited intensity resolutionFrame buffers have limited number of bits per pixelPhysical devices have limited dynamic range

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

255 150 75 0

Blue channel

Green channel

Red channel

KUCG


Image Display

Re-create continuous function from samplesExample: cathode ray tube

Image is reconstructed by displaying pixels with finite area (Gaussian)

KUCG


Image Resolution

Intensity resolutionEach pixel has only “Depth” bits for colors/intensities

Spatial resolutionImage has only “Width” x “Height” pixels

Temporal resolutionMonitor refreshes images at only “Rate” Hz

KUCG


Overview

Image RepresentationWhat is an image?

Sampling and ReconstructionKey steps in image processing

KUCG


Sampling and Reconstruction

Sampling

Reconstruction

KUCG


Sampling and Reconstruction

KUCG


Image Processing

Pixel operationsAdd luminanceAdd contrast

FilteringBlurDetect edge

WarpingScaleRotateWarps

KUCG


Adjusting Brightness

Simply scale pixel componentsMust clamp to range (e.g., 0 to 255)

Original Brighter

KUCG


Adjusting Contrast

Compute mean luminance L for all pixelsLuminance = 0.30*r + 0.59*g + 0.11*b

Scale deviation from L for each pixel componentMust clamp to range (e.g. 0 to 255)

Original Contrast

L

KUCG


Image Processing




KUCG


Adjusting Blurriness

Convolve with a filter whose entries sum to oneEach pixel becomes a weighted average of its neighbors

Original Blur

⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢

⎣

⎡

=

161

162

161

162

164

162

161

162

161

Filter

KUCG


Edge Detection

Convolve with a filter that finds differences between neighbor pixels

Original Edge Detection

⎥⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡

−−−−−−−−

=111181111

Filter

KUCG


Image Processing




KUCG


Image Warping

Move pixels of imageMappingResampling

Warp

Source Image Destination Image

KUCG


Overview

MappingForwardReverse

ResamplingPoint samplingTriangle filterGaussian filter

KUCG


Mapping

Define transformationDescribe the destination (x, y) for every location (u, v) in the source (or vice-versa, if invertible)

u

v

x

y

KUCG


Example Mappings

Scale by factor :x = factor * uy = factor * v

Scale0.8

u

v

x

y

KUCG


Example Mappings

Rotate by θ degrees:x = u cos θ – v sin θy = u sin θ + v cos θ

u

v

Rotate30

x

y

KUCG


Example Mappings

Shear in X by factor :x = u + factor * vy = v

Shear in Y by factor :x = uy = v + factor * u

Shear X1.3

u

v

x

y

Shear Y1.3

u

v

x

y

KUCG


Other Mappings

Any function of u and v :x = fx(u, v)y = fy(u, v)

Fish-eye Swirl Rain

KUCG


Image Warping Implementation I

Forward mapping :for(int u=0; u<umax; u++) {

for(int v=0; v<vmax; v++) {float x = fx(u,v);float y = fy(u,v);dst(x,y) = src(u,v);

}}


(u, v)(x, y)f

KUCG


Forwarding Mapping

Iterate over source image

Rotate-30

u

v

x

y

KUCG


Forwarding Mapping – NOT


Rotate-30

u

v

x

y

Many source pixels can map to same destination pixel

KUCG


Forwarding Mapping – NOT


Rotate-30

u

v

x

y

Many source pixels can map to same destination pixel

Some destination pixels may not be covered

KUCG


Image Warping Implementation II

Reverse mappingfor(int x=0; x<xmax; x++) {

for(int y=0; y<ymax; y++) {float u = fx

-1(x,y);float v = fy

-1(x,y);dst(x,y) = src(u,v);

}}


(u, v)(x, y)f

KUCG


Reverse Mapping

Iterate over destination imageMust resample sourceMay oversample, but much simpler!

Rotate-30

u

v

x

y

KUCG


Resampling

Evaluate source image at arbitrary (u, v)

(u, v) does not usually have integer coordinates


(u, v)(x, y)

KUCG


Overview

MappingForwardReverse

ResamplingPoint samplingTriangle filterGaussian filter

KUCG


Point Sampling

Take value at closest pixelint iu = trunc(u+0.5);int iv = trunc(v+0.5);dst(x, y) = src(iu, iv);

This method is simple, but it causes aliasing

Rotate-30

Scale0.5

u

v

x

y

KUCG


Triangle Filtering

Convolve with triangle filter

Input Output

KUCG


Triangle Filtering

Bilinearly interpolate four closest pixelsa = linear interpolation of src(u1, v2) and src(u2, v2)b = linear interpolation of src(u1, v1) and src(u2, v1)dst(x, y) = linear interpolation of “a” and “b”

(u1, v2)

(u1, v1)

(u2, v2)

(u2, v1)

a

b

(u, v)

KUCG


Gaussian Filtering

Convolve with Gaussian filter

Width of Gaussian kernel affects bluriness

Input Output

KUCG


Gaussian Filtering

Compute weighted sum of pixel neighborhood :Weights are normalized values of Gaussian function

(u, v)

KUCG


Filtering Methods Comparison

Trade-offsAliasing versus blurringComputation speed

Point Bilinear Gaussian

KUCG


Image Warping Implementation III




-1(x,y);dst(x,y) = resample_src(u,v,w);

}}


(u, v)(x, y)f

KUCG


Image Warping Implementation III




-1(x,y);dst(x,y) = resample_src(u,v,w);

}}


(u, v)(x, y)f

w

KUCG


Example: Scale

Scale (src, dst, sx, sy) :float w ≈ max(1/sx, 1/sy)for(int x=0; x<xmax; x++) {

for(int y=0; y<ymax; y++) {float u = x/sx ;float v = y/sy;dst(x,y) = resample_src(u,v,w);

}}

Scale0.5

u

v

x

y

(u, v)

(x, y)

KUCG


Example: Rotate

Rotate (src, dst, theta)for(int x=0; x<xmax; x++) {

for(int y=0; y<ymax; y++) {float u = x*cos(-θ)-y*sin(-θ)float v = x*sin(-θ)+y*cos(-θ)dst(x,y) = resample_src(u,v,w);

}}

u

v

Rotate30

x

y

(u, v) (x, y)

KUCG


Example: Fun

Swirl (src, dst, theta)for(int x=0; x<xmax; x++) {

for(int y=0; y<ymax; y++) {float u = rot(dist(x,xcenter)*θ)float v = rot(dist(y,ycenter)*θ)dst(x,y) = resample_src(u,v,w);

}}

u

v

Swirl45

x

y(u, v) (x, y)

Documents

Image Quantization, Halftoning, and Ditheringkucg.korea.ac.kr/education/2008_2/CNCE340/tutor_/08ImageProcessing.pdfGraphics Lab @ Korea University Image Resolution Intensity resolution