wk 8.1 Flashcards
(19 cards)
what is image compression
method of data compression on digital image
objectives of image compression
- reduce data/ bits to represent an image
- store and transmit data efficiently
Data vs Information
- Different amounts of data can be used to
represent the same amt of info - Representations that contain repeated/irrelevant
info is said to contain redundant data
how is compression possible
Compression Algorithms take advantage of the difference between data & information inherent in image data
types of data redundancies
code, spatial, psychovisual
what is Code Redundancy
when data representing image is not utilised optimally
what is code in context of Code Redundancy
Code is a system of symbols representing body of
information / piece of info = 1 code symbol
In images: 8-bit code (1 pixel) used to represent grayscale value/intensity. However, may contain more bits than required to represent intensity
Spatial Redundancy
information pixels replicate unnecessarily
Pixels of most 2D array images are correlated spatially ie. adjacent pixels brightness levels do not change rapidly, change is gradual.
Psychovisual Redundancy
Most 2D intensity arrays contain information that is ignored by human vision system.
considered redundant as its not used.
compression ratio D2 D1
1st representation of the image (D1)
2nd representation of the image (D2)
D1 -> compression -> D2
D2 = D1
1st representation of the image (D1) is does not contain any redundant data
D2 less than D1
High compression ratio
1st representation of the image (D1) contains highly redundant data & there is significant compression to D1
D2 more than D1
Low Compression Ratio
2nd representation of the image (D2) contains more data than D1
If an uncompressed image is file size 10MB and is compressed to 2MB
compression ratio = 10 : 2
= 5
If an uncompressed image has 8-bit per pixel & is compressed such that each pixel is now
represented by only 1 bit per pixel
compression ratio = 8:1
= 8
lossless data compression methods
text/programs
bit plane, run length, huffman, lempel ziv
lossy data compression methods
image/video/audio
jpeg, mpeg, mp3
what is Lossless Image Compression
compressed file can be decoded back to its original form without information loss
- Original can be recovered exactly which is high quality but bigger
- Preferred form of compression, but has low compression ratios
uses of Lossless Image Compression
images generally for some medical/legal/scientific/political significance
◦ Binary images/black-and-white graphics are good candidates for lossless compression
◦ Some colour & grayscale images that contain series of identical pixel may also be used
