Basic definitionof image compression
Data redundancy
Elements of information theory
General mechanism and types of data
compression and image restoration
Huffman coding
Arithmetic coding
Dictionary based coding
Bit-plane coding
Outlines
3.
⦿ It isthe art and science of reducing the
amount of data required to represent an
image.
⦿ Image compression is the process of reducing
the size of an image file without significantly
degrading its visual quality.
⦿ This is achieved by removing redundant or
unnecessary data from the image.
Image compression
4.
⦿ Driving factorsbehind image compression
Storage space
Shorten transmission time
⦿ It is essential for saving storage space,
reducing bandwidth usage, and improving
load times for websites and applications.
⦿ To achieve Compression , Reduction of
redundant data is necessary.
Original
Image compress
Compressed
Image file decompress extracted
Image file
Con’t
5.
⦿ Compression techniquesare broadly classified as
Lossless and Lossy.
⦿ Lossless Compression
◾ Reduces file size without losing any data.
◾ The original data can be perfectly reconstructed from the
compressed version.
◾ It is used mainly for compressing texts, executable programs,
spreadsheets, etc..
◾ Medical imaging , Space images ,Technical drawings etc
⦿ Lossy Compression
◾ Reduces file size of an image by permanently eliminating certain
amounts of data, particularly details that are less noticeable to the
human eye.
◾ This compression is best suited for images where reduction in
file size is more critical than preserving the exact original
quality.
◾ Advantage is higher compression
◾ Web images, Photographic images
Compression techniques
⦿ Huffman codingis the most popular variable
length coding technique.
⦿ It is based on the frequency of occurrence
of data items and
⦿ is particularly effective when some data items
occur much more frequently than others.
⦿ Forward Pass
1. Sort probabilities per symbol
2. Combine the lowest two probabilities
3. Repeat Step2 until only two probabilities remain.
8.
⦿ Arithmetic codingis a more advanced
lossless compression technique.
⦿ Arithmetic coding represents a sequence of
symbols(such as pixel values) as a single
number in range [0,1]. This results in a
more compact representation.
⦿ Arithmetic coding yields better
compression
⦿ Entire sequence of source symbols is assigned a
single arithmetic code word
⦿ Disadvantage
Slower than Huffman coding
Random access is difficult
9.
⦿ Dictionary-based codingis a powerful method
for lossless data compression that relies on
identifying and encoding repeated patterns in
the input data.
⦿ Unlike statistical methods such as
Huffman and arithmetic coding, which
focus on individual symbol probabilities,
dictionary-based methods compress data
by finding and encoding repeated patterns
or substrings.
⦿ The most well known dictionary-based
algorithm is the Lempel-Ziv-Welch(LZW)
algorithm.
Dictionary-based coding
10.
⦿ Bit-plane codingis a technique used
primarily in image compression, where
the image is decomposed into a series of
binary images (bit-planes).
⦿ Each bit-plane represents a single bit
position across all pixel values in the
image.
⦿ This method can be used for both lossless
and lossy compression.
Bit-plane coding
11.
⦿ Discrete CosineTransform (DCT): Used in
JPEG compression, it transforms the image into
a sum of cosine functions oscillating at
different frequencies.
⦿ Quantization: This process reduces the
number of colors or brightness levels in an
image, which reduces the amount of data
needed to represent the image.
⦿ JPEG (Joint Photographic Experts Group):
The most widely used lossy compression
format.
⦿ It uses DCT, quantization, and entropy coding.
Common lossy compression
techniques
12.
⦿ Lossless Compression:Best for images where
quality is paramount, such as medical imaging,
technical drawings, and images that require
further editing. (e.g., PNG)
⦿ Lossy Compression: Suitable for everyday
photos, web images, and scenarios where a
balance between quality and file size is needed.
(e.g., JPEG or WebP)
Choosing the Right Compression
13.
⦿ Telecommunications: Efficientdata
transmission over limited bandwidth channels.
⦿ Digital Storage: Reduced storage requirements
for digital files.
⦿ Medical Imaging: Restoration techniques are
crucial for enhancing the quality of medical
images for accurate diagnosis.
⦿ Photography and Videography: Compression
helps in managing large image and video files,
while restoration can improve the quality of old
or degraded media.
Applications of Compression and Restoration
14.
⦿ Adobe Photoshop:Provides options for saving
images in various formats with different compression
settings.
⦿ GIMP(GNU Image Manipulation Program): is a free
and open source raster graphics editor used for
image retouching, editing, composition and more.
⦿ An open-source alternative to Photoshop with
support for various compression techniques.
⦿ ImageMagick: A command-line tool for image
manipulation and compression.
⦿ Online Services: Websites like TinyPNG and JPEG-
Optimizer offer convenient image compression
services.
Tools and Software
15.
⦿ The useof color is important in image
processing because:
⦿ Color is a powerful descriptor that simplifies
object identification and extraction.
⦿ Humans can discern thousands of color shades
and intensities, compared to about only two
dozen shades of gray..
Color Image Processing
16.
⦿ Colors areseen as variable
combinations of the primary colors of
light:
⦿ primary colors of light: red (R), green
(G), and blue (B).
⦿ The primary colors can be mixed to
produce the secondary colors:
⦿ secondary colors: magenta (red+blue),
cyan (green+blue),and yellow
(red+green).
⦿ Mixing the three primaries, or a
secondary with its opposite primary
color, produces white light.
Color Fundamentals
Figure 1: Primary and
secondary colors of light
17.
⦿ Color imageprocessing involves the
manipulation and analysis of images that
contain color information, as opposed to
grayscale images which only contain intensity
information.
Color image processing
18.
The organizationof the colors of in an image in a
specific format is called color space.
The way in which a color is represented is called a
color model.
Each and every image uses one of the following color
spaces for effective picture representation:
RGB (Red, Green, Blue):
⦿ The most common color space where colors are
represented as combinations of red, green, and blue
intensities.
HSV (Hue, Saturation, Value):
⦿ Represents colors in terms of hue (color type),
saturation (intensity of the color), and value
(brightness).
Color space
19.
HSL (Hue, Saturation,Lightness):
⦿ Similar to HSV but with lightness instead of
value.
CMYK (Cyan, Magenta, Yellow, Key/Black):
⦿ Used in printing, representing colors as
combinations of cyan, magenta, yellow, and
black.
YUV/YIQ:
⦿ Used in video systems, separating luminance
(Y) from chrominance (UV or IQ).
Con’t
20.
⦿ Medical Imaging:Enhancing and analyzing
medical images for diagnosis.
⦿ Remote Sensing: Processing satellite or aerial
images for environmental monitoring.
⦿ Computer Vision: Enabling machines to
interpret and understand visual information.
⦿ Digital Photography: Enhancing and editing
photos for better visual appeal.
Applications of Color Image Processing
21.
⦿ OpenCV: Apowerful library for image
processing tasks, including color image
processing.
⦿ MATLAB: Provides extensive tools for image
processing and analysis.
⦿ Python Libraries: Such as PIL/Pillow, scikit-
image, and NumPy for image manipulation.
Tools and Libraries