Image compression

Question 1

What is 100 photo quality?

Answer 1

Perfect image quality with no loss of information

Question 2

When the compression ratio is high…

Answer 2

The image ratio is low

Question 3

What is non-lossy compression?

Answer 3

• Non-lossy (lossless) compression – original image can be recovered exactly from compressed image file.
• E.g. Portable Network Graphics (.png)

Question 4

What is lossy compression?

Answer 4

• Lossy compression – Information lost when file is compressed. Therefore only an approximation to the original image is possible.
• Eg. Joint Experts Photographic Group JPEG (.jpg)

Question 5

What are the 3 types of data redundancy?

Answer 5

• Coding redundancy
• Inter-pixel redundancy
• Psycho-visual redundancy

Question 6

What causes coding redundancy?

Answer 6

It is caused by sub-optimal code words for symbol encoding - a symbol typically (but not exclusively) represents grey-level.

Question 7

What causes inter-pixel redundancy?

Answer 7

Due to grey-level correlations between neighbouring pixels.

Question 8

What is psycho-visual redundancy?

Answer 8

Information contained within an image that is superfluous to the interpretation or aesthetics of an image.

Question 9

What does clarity depend on?

Answer 9

• Spatial frequency
• Amplitude

Question 10

What is the compression ratio?

Answer 10

A measure of the the compactness of a compressed image file

Question 11

How is compression ratio (Cr) defined?

Answer 11

CR = b1 / b2
b1 = the number of bits in the original image
b2 = the number of bits in the encoded image

Question 12

What are the steps from A to the compressed image?

Answer 12

• Transformation
• Quantization
• Symbol encoder

Question 13

What are the steps from compressed image to A?

Answer 13

• Symbol decoder
• Dequantizer
• Inverse transformation

Question 14

What does symbol refer to?

Answer 14

Symbol is a general term that can refer to pixel values or transformation coefficient values.

Question 15

What happens in the transformation step?

Answer 15

• Converts the input image A into a format that is better suited to encoding.
• This transforms the pixels into coefficients.
• For most natural images a significant number of transformation coefficients have negligible magnitude and can be discarded.
• The remaining coefficients can be coarsely quantized – thereby reducing the number of bits required to encode them.

Question 16

What happens in the quantization step?

Answer 16

• The quantizer achieves compression by approximating the transformation coefficients.
• Quantization is an inherently lossy process whereas the image transform and coding processes are not.
• Therefore a non-lossy compression method can not comprise a quantization component.

Question 17

What happens in the symbol coding step?

Answer 17

Symbol coding compresses the image by exploiting the fact that in natural images some grey-levels occur more frequently than others (also applies to transformation coefficients).

Question 18

What is the JPEG image compression recipe?

Answer 18

• Split into 8x8 blocks and treat each block separately.
• Apply Discrete Cosine Transformation (DCT) to each image block.
• Quantise the DCT coefficients (transformed 8x8 pixel block) .
• Apply Huffman coding scheme (symbol coder) to quantised coefficients.

Question 19

In JPEG image compression transformations where are the significant coefficients found?

Answer 19

The significant coefficients of the DCT are usually located in the top left-hand corner

Question 20

What does a quantizer do and how is it achieved?

Answer 20

• A quantizer reduces the number of bits needed to store the coefficients that result from an image transformation.
• This is achieved by reducing the precision of the coefficients.

Question 21

What is cosine referred to?

Answer 21

The basis function

Question 22

What does weighting refer to?

Answer 22

Coefficient and relates to pixel values