DIGITAL IMAGE

PROCESSING
UNIVERSITY OF ENGINEERING & TECHNOLOGY,
MARDAN
ASSIGNMENT

• MATLAB Codes
• Log/Anti-log
• Power-law transformations
• Inputs for r, gamma, c from user
• M file for each transformation
• Graph of s v/s r
• Hard copy & soft copy to CR, working m file
• If copied/plagiarised, penalty
PIECEWISE LINEAR
TRANSFORMATION FUNCTIONS
• OTHER THAN LINEAR, LOG & POWER LAW
TRANSFORMATIONS, WE CAN HAVE PIECEWISE
LINEAR TRANSFORMATIONS
• PRACTICAL IMPLEMENTATION OF MANY
TRANSFORMATIONS CAN BE MODELLED AS
PIECEWISE LINEAR
• DISADVANTAGE: MORE USER INPUT REQUIRED
(BY COMPARISON)
QUESTION

• WHY WOULD WE HAVE A LOW CONTRAST IMAGE

• POOR ILLUMINATION
• LACK OF DYNAMIC RANGE IN IMAGING SENSOR
• WRONG SETTING OF LENS APERTURE
PIECEWISE LINEAR
TRANSFORMATION FUNCTIONS
• CONTRAST STRETCHING
• TO INCREASE THE DYNAMIC RANGE OF THE GRAY LEVELS IN THE IMAGE BEING
PROCESSED
• LOCATIONS OF (s1,r1) & (s2,r2) CONTROL THE SHAPE OF TRANSFORMATION
• IF s1=r1, s2=r2,.....
• IF s1=0,s2=L-1,r1=r2,....... (b) in FIGURE
• INTERMEDIATE VALUES OF (r1,s1) & (r2,s2) PRODUCE VARYING DEGREES OF STRETCH
IN GRAY LEVELS
• GENERALLY, r1<=r2 & s1<=s2, FUNCTION IS MONOTONICALLY INCREASING & SINGLE
VALUED, PRESERVING THE ORDER OF GRAY LEVELS
• A. PWLF TRANSFORMATION GRAPH
• B. 8-BIT LOW CONTRAST IMAGE
• C. (r1,s1)=(rmin,0), (r2,s2)=(rmax,L-1)
• Stretch from (rmin,rmax) to (0,L-1)
• D. r1=r2=m (mean gray level in the image), (s1,s2)=(0,L-1)
CONTRAST STRETCHING (PWLF)
RECALL CONTRAST
STRETCHING
PIECEWISE LINEAR
TRANSFORMATION FUNCTIONS
• GRAY-LEVEL SLICING
• HIGHLIGHTING A SPECIFIC RANGE OF GRAY LEVELS IN AN
IMAGE
• MASSES OF WATER IN SATELLITE IMAGES
• FLAWS IN AN XRAY IMAGES
• 2 APPROACHES
• DISPLAY HIGH VALUE FOR ALL GRAY LEVELS IN RANGE
OF INTEREST; LOW VALUE FOR OTHER GRAY LEVELS (A)
• BRIGHTEN DESIRED RANGE OF GRAY LEVELS BUT
PRESERVE THE BACKGROUND (B)
GRAY-LEVEL SLICING
PIECEWISE LINEAR
TRANSFORMATION FUNCTIONS
• BIT-PLANE SLICING
• HIGHLIGHTING THE CONTRIBUTION BY SPECIFIC BITS
• SUPPOSE EACH PIXEL IS REPRESENTED BY 8-BITS
• IMAGINE THAT IMAGE IS COMPOSED OF 8 1-BIT
PLANES
• BIT-PLANE 0 FOR LSB, BIT PLANE 7 FOR MSB
• PLANE 0 CONTAINS ALL THE LOWEST ORDER BITS
• PLANE 7 CONTAINS ALL THE HIGHEST ORDER BITS
 BIT-PLANE SLICING

• SEPARATING A DIGITAL IMAGE INTO ITS BIT PLANES IS USEFUL

FOR ANALYZING THE RELATIVE IMPORTANCE OF EACH BIT OF
IMAGE
• USEFUL FOR IMAGE COMPRESSION
• USED FOR DETERMINING THE ADEQUACY OF THE NUMBER OF BITS USED
TO QUANTIZE EACH PIXEL
BIT-PLANE SLICING

• Digitally, an image is represented in terms of pixels

• These pixels can be expressed further in terms of bits.
• Consider the image ‘coins.png’ and the pixel representation
of the image.
BIT PLANE SLICING

• Consider the pixels that are bounded within the yellow line. The binary formats for
those values are (8-bit representation)

• The binary format for the pixel value 167 is 10100111

• Similarly, for 144 it is 10010000
• This 8-bit image is composed of eight 1-bit planes.
BIT PLANE SLICING

• Plane 1 contains the lowest order bit of all the pixels in the image.

• And plane 8 contains the highest order bit of all the pixels in the
image.
BIT-PLANE SLICING
 BIT-PLANE SLICING
HIGHER ORDER BITS CONTAIN MAJORITY OF
VISUALLY SIGNIFICANT DATA
OTHER BIT PLANES CONTRIBUTE TO MORE
SUBTLE DETAILS IN THE IMAGE
BIT-PLANE SLICING
BIT-PLANE SLICING
QUESTION

• APPLY A THRESHOLDING GRAY-LEVEL

TRANSFORMATION FUNCTION TO 256 LEVEL
GRAYSCALE IMAGE
• MAPS ALL VALUES BETWEEN 0 & 127 TO ONE LEVEL
• ALL VALUES FROM 128 TO 255 MAPPED TO OTHER LEVEL