Digital Image Processing

Digital Image Processing Hongjun Xu August 1, 2007

Lecture 1 Introduction and Digital Image Fundamentals

• What is an Image? Picture, photograph Visual data Usually two or three dimensional

What is a digital image? An image which is “discretized,”, i.e., defined on a discrete grid Two-dimensional collection of light values (or gray values)

What is digital image processing? Digital image processing is the study of representation and manipulation of pictorial information by a computer. Improve pictorial information for better clarity (human interpretation) Examples: 1 Enhancing the edges of an image to make it appear sharper 2 Remove “noise” from an image 3 Remove motion blur from an image

Automatic machine processing of scene data (interpretation by a machine/non-human, storage, transmission) Examples: 1 Obtain the edges of an image 2 Remove detail from an image

What is image interpretation? Assign meaning to an ensemble of recognized objects

Important steps in a typical image processing system Step 1: Image acquisition Capturing visual data by an image sensor Step 2: Discretization / digitalization; Quantization ; Compression Convert data into discrete form; compress for efficient storage/transmission Step 3: Image enhancement and restoration Improving image quality (low contrast, blur noise) Step 4: Image segmentation Partition image into objects or constituent parts Step 5: Feature selection Extracting pertinent features from an image that are important for differentiating one class of objects from another

Step 6: Image representation Assigning labels to an object based on information provided by descriptors Step 7: Image interpretation Assigning meaning to an ensemble of recognized objects

Imaging:

Step 1: Image acquisition and Step 2: digitalization; Quantization ; Compression

Image processing: Step 2: Compression Step 3: Image enhancement and restoration Step 4: Image segmentation Image analysis: Step 4: Image segmentation Step 5: Feature selection Step 6: Image representation Image understanding: Step 6: Image representation Step 7: Image interpretation

Image processing and analysis transformations Level 0: Image representation Acquisition; sampling; quantization; compression Level 1: Image to image transformations Enhancement; restoration; segmentation Level 2: Image to parameter transformation Feature selection Level 3: Image to decision transformation Recognition and interpretation

Image Acquisition and Sampling Sampling refers to the process of digitizing a continuous function 1 For Example: y  sin( x)  sin(3 x) 3

Sampling an image requires that we consider the Nyquist criterion, when we consider an image as a continuous function of two variables, we wish to sample it to produce a digital image. Using light Light is the predominant energy source for image because it is the energy source which human beings can observe directly. For example, photographs are pictorial record of visible scene. Many digital images are captured using visible light as the energy source.

Light spectrum

Types of sensors: • Optical (camera) • Infrared (senses heat changes) • X-ray (CT Scan) • Magnetic (MRI) • Ultrasound (acoustic energy) • Electron Microscopy (Electron beam) • Computer generated images (fractals, animation)

Create an image To create a digital image, we need to convert the continuously sensed data into digital form. This involves: Sampling: Digitizing the coordinate values (resolution) Depends on density of sensor in an array Limited by optical resolution Quantization (bits/pixel) Digitizing the amplitude values Pixel: short for picture element

Image Representation

Spatial Resolution Sampling determines resolution Resolution is the smallest discernible detail in an image Common unit – 480 x 640 pixels Need to know size of image also

Quantization Gray: generally integer values, ranging from 0 to some maximum value

• 1 bit /pixel –2 possible values –2 gray levels -> 0 or 1 (binary image)

• B bits/pixel –2B gray levels –1 byte = 8 bits –> 256 levels

Coordinate Convention

Digital Image Representation  f (0,0)  f (1,0) f ( x, y ) =   ...   f ( N − 1,0)

Digital Image

f (0,1)

...

...

...

... ... f ( N − 1,1) ...

f (0, M − 1)  f (1, M − 1)   ...  f ( N − 1, M − 1)

Image Elements (Pixels)

Pixel Notation

Reading Images MATLAB syntax: imread(‘filename’) Format name

Description

Recognized Extensions

TIFF

Tagged Image File Format

.tif, .tiff

JPEG

Joint photographic experts group

.jpg, .jpeg

GIF

Graphics Interchange format

.gif

BMP

Windows Bitmap

.bmp

PNG

Portable Network Graphics

.png

XWD

X Window Dump

.xwd

Note: GIF is supported by imread, but not by imwrite.

Applications • BIOLOGICAL: automated systems for analysis of samples. •DEFENSE/INTELLIGENCE: enhancement and interpretation of images to find and track targets. •DOCUMENT PROCESSING: scanning, archiving, transmission. •FACTORY AUTOMATION: visual inspection of products. •LAW ENFORCEMENT/FORENSICS: fingerprint analysis. •MATERIALS TESTING: detection and quantification of cracks, impurities, etc. •MEDICAL: disease detection and monitoring, therapy/surgery planning

•...

Image processing with MATLAB What is the Image Processing Toolbox? • The Image Processing Toolbox is a collection of functions that extend the capability of the MATLAB ® numeric computing environment. The toolbox supports a wide range of image processing operations, including: – – – – – – –

Geometric operations Neighborhood and block operations Linear filtering and filter design Transforms Image analysis and enhancement Binary image operations Region of interest operations (RIO)

>> help images

Data Classes Double

-double precision, range: [-10^308, +10^308]

Uint8

-unsigned 8-bit integer, range:[0,255]

Unit16

-unsigned 16-bit integer, range:[0, 65535]

Unit32

-unsigned 32-bit integer, range:[0, 2^32-1]

Int8

-signed 8-bit integer, range:[-128,127]

Int16 -signed 16-bit integer, range:[-32768,+32767] Int32 Single

-single precision, range:[-10^38,+10^38]

Char

-characters

Logical

-values are 0 or 1

Converting between Image Classes Name

Convert input to

Im2uint

uint8

Im2uint16

uint16

Im2double

double

Im2bw

logical

Mat2gray

double (in range [0,1])

Image types in MATLAB 1. White and black images They are also called binary images, containing 1 for white and 0 for black. 2. Grey scale images They are also called Intensity Images, containing numbers in the range of 0 to 255 or 0 to 1. 3. Colored image They may be represented as RGB Image or Indexed Image.

RGB images: m-by-n-by-3 If each of these components has a range 0-255, this gives a total 255^3=16777216 different possible colours in the image Indexed images: m-by-3 color map Most colour images only have a small subset of the more than sixteen million possible colours. So the image has an associated colour map. Each pixel has a value which does not give its colour, but an index to the colour in the map.

Gray-level values can be stored as one of several data types: –uint8: 8-bit unsigned integer; range is 0 (black) - 255 (white) –double: double-precision floating-point number; range is 0.0 - 1.0 –uint16 All MATLAB operations work on images of type double Only some MATLAB operations work on intensity or binary images of type uint8 >>x=double(x); DOUBLE Convert to double precision.

Binary images MATLAB code

Intensity Images

RGB images

Indexed Images

MATLAB stores the RGB values of an indexed image as values of type double, with values between 0 and 1.

Image display • • • • • • • •

image - create and display image object imagesc - scale and display as image imshow - display image colorbar - display colorbar getimage- get image data from axes truesize - adjust display size of image zoom - zoom in and zoom out of 2D plot Imread - read image

Writing Images imwrite

-write images

Basic syntax: imwrite (f,’filename’)

Image conversion

• • • • • • • • •

gray2ind - intensity image to index image im2bw - image to binary im2double - image to double precision im2uint8 - image to 8-bit unsigned integers im2uint16 - image to 16-bit unsigned integers ind2gray - indexed image to intensity image mat2gray - matrix to intensity image rgb2gray - RGB image to grayscale rgb2ind - RGB image to indexed image