1. Dr. Jim Rowan ITEC 2110 Bitmapped Images
Digital Media
Dr. Jim Rowan
ITEC 2110
Bitmapped Images

2. Device Resolution
Determines how finely the device approximates the continuous phenomenon
Is closely related to sampling we discussed earlier
Can be expressed a number of different ways
Printers and Scanners?
Number of dots per inch
Number of pixels per inch
Video?
Number of pixels, pixel x and y dimensions

3. Device Resolution
When considering scanners and printers pay attention to the resolution.
The number of dots per inch a printer produces will dictate the print size of the image
This can cause what appears to be a small image to become quite large

4. Device Resolution vs Printed Size
Bigger than you thought?
If the printer has a 72 dpi rating and the image was scanned at 600 dpi, printing the image (unscaled) will result in a large image 600/72 = 8.33 times as large
To scale it to get the original size back you would use a scaling factor of 72/600 or 0.12

5. Device Resolution vs Printed Size
Smaller than you thought?

6. Device Resolution vs Printed Size
Smaller than you thought?

8. Image Formats
The pixel x and y dimensions of the image can be seen as a measure of how much DETAIL is contained in the picture
Most encode (put in the header) the resolution of the image in Pixels Per Inch (PPI)
Many encode (put in the header) the original size as pixel width and pixel height

9. Resolution Changes? Resolution Increase…
Is image resolution lower than the the output device?
Must scale it up...
Must add pixels...
Requires interpolation between pixels
Always results in an APPARENT quality REDUCTION in the image

10. Here the original 4x4 image
is doubled in size to
8x8 by adding pixels

11. If you double the image size you have to add pixels...
But what color do you make
the additions? ?

12. Generally you consider what the colors are that surround
the original pixel
Mathematically this usually
takes the form of matrix operation ?

13. Resolution Decrease…
Is image resolution higher than the output device?
Must discard some pixels...
AKA downsampling
Downsampling: A paradox
There are fewer bits since you’re throwing some pixels out
But... subjective quality goes up
How? Downsampling routine can use the tossed-out pixels to modify the remaining pixel
Intentionally doing this is called oversampling
How to do this? ==>

14. If you cut the image size in half
(8x8 -> 4x4)-> = 48
pixels removed
You remove 3/4 of the pixels!
What do you do with thrown
away pixels?
64 pixels
16 pixels

15. One answer: throw them away!
Here it works...
because it is a solid color

16. Another answer:
Use the information
in the surrounding pixels to
influence the remaining pixel

17. Convolution Calculations
Convolution is the mathematical process that image software (like GIMP or Photoshop) use to do special effects More at the end of this lecture…

18. Browsers... really bad at downsampling
Their image processing is not very sophisticated
What are the implications?
Use image processing programs to do downsampling
(GIMP, Photoshop) are sophisticated enough to take advantage of the extra information so...
Images for WWW should be downsampled before they are used on the web.

19. Data Compression What we’ve seen so far: How to reduce that?
Storing an image as an array of pixels
With color stored as three bytes per pixel
Image file gets BIG fast!
How to reduce that?
Using a color table reduces the file size of the stored image
Other data compression techniques ==>

20. Data Compression Consider this image: With no compression...
RGB encoding =>
64 x 3 = 192 bytes
64 pixels

21. Data Compression Run Length Encoding
Consider this image:
RLE compression...
9RGB6RGB2RGB6RGB2RGB6RGB2RGB6RGB2RGB6RGB2RGB6RGB9RGB
= 49 bytes
64 pixels

22. Run Length Encoding
This advantage would be dependent on the CONTENT of the image.
Why?
Could it result in a larger image?
How?
Generally, any data compression CAN result in a larger file than using the pixel array storage
Dependent on the image contents

23. Run Length Encoding: Always better than RGB?
Consider this image:
RLE compression...
1RGB1RGB1RGB1RGB1RGB... 1RGB1RGB1RGB
-> 256 bytes
(a tiny lie!)
RGBRGBRGB... RGBRGB
-> 192 bytes
64 pixels

24. Run Length Encoding RLE is Lossless What is lossless? Original Exact
compression routine
Original
compressed
original
decompress routine
Exact
duplicate
Original

25. Dictionary-based (aka Table-based) compression technique
(Note: Data compression works on files other than images)
Construct a table of strings (colors) found in the file to be compressed
Each occurrence in the file of a string (color) found in the table is replaced by a pointer to that occurrence.

26. Data Compression Dictionary-based(Table-based) We’ve seen this!
Consider this image:
RGBRGB==>
[ ][ ]
[ ].[ ]
[ ][ ]
[ ][ ]
[ ][ ]
[ ][ ]
->14 bytes
64 pixels

27. Lossless techniques Can be used on image files
Lossy techniques toss some data out
-jpeg is a lossy technique
Must be used for executable files
Why?

28. JPEG compression
Best suited for natural photographs and similar images
Fine details with continuous tone changes
High frequency components are associated with abrupt changes in image intensity
JPEG takes advantage that humans don’t perceive the effect of high frequencies accurately

29. JPEG compression... JPEG finds these high frequency components by
treating the image as a matrix
using the Discrete Cosine Transform (DCT) to convert an array of pixels into an array of coefficients
DCT is expensive computationally so it the image is broken into 8x8 pixel squares and applied to each of the squares

30. JPEG compression... DCT does not actually compress the image
Allows most of the high frequency components to be discarded because they do not contribute much to the perceptible quality of the image
Encodes the frequencies at different quantization levels giving the low frequency components more quantization levels
==>JPEG uses more storage space for the more visible elements of an image

31. JPEG compression... Lossy
Effective for the kinds of images it is intended for ==> 95% reduction in size
Allows the control of degree of compression
Suffers from artifacts that causes edges to blur... WHY?
HMMMmmmm…

32. Side Note! To make matters worse…
The human vision system is very complex
Upside down
Split- left side of eye to right side of brain
Right side of eye to left side of brain
Cones and rods not uniformly distributed
Cones and rods are upside down resulting in blind spots in each eye that we just ignore!
Partially responsible for making lossy techniques to work ==>

33. One reason lossy compression works

34. Optical Illusions
See Additional Class Information: Illusions

35. Image Manipulation with GIMP
Why?
Correct deficiencies (i.e. flash red eye)
encapsulated sequence of operations to perform a particular change
Create images that are difficult or impossible to create in nature
special effects

36. Pixel Point Processing
Allows adjustment of color in an image
Color adjustment, linear
brightness
adjusts every pixel brightness up or down
contrast
adjusts the RANGE of brightness
increasing or reducing the difference between brightest and darkest areas

37. Remember this dilemma?
Rescaling a bitmapped image is called resampling: Two kinds
Downsampling
Upsampling
Different ways to do this that result in different results
P111 Nearest Neighbor, bilinear a& bicubic

39. Pixel Group Processing
Final value for a pixel is affected by its neighbors
Because the relationship between a pixel and its neighbors provides information about how color or brightness is changing in that region
How do you do this?
==> Convolution!

40. Convolution & Convolution Masks
Very expensive computationally
each pixel undergoes many arithmetic operations
If you want all the surrounding pixels to equally affect the pixel in question... use a evenly weighted convolution mask

41. 1/9 X
Convolution
mask
Convolution
kernel
Using this convolution mask
on this convolution kernel
the final value of the pixel (2,2)
will be:
pixel (2,2) = 1/9(1,1) + 1/9(1,2)+ 1/9(1,3)
+1/9(2,1) +1/9(2,2) +1/9(2,3)
+1/9(3,1) +1/9(3,2) +1/9(3,3) X

42. 1/9 X
Convolution
mask
Using this convolution mask
on this convolution kernel
the final value of the pixel (3,2)
will be:
pixel (3,2) = 1/9(1,2) + 1/9(1,3)+ 1/9(1,4)
+1/9(2,2) +1/9(2,3) +1/9(2,4)
+1/9(3,2) +1/9(3,3) +1/9(3,4) X

43. 1/9 X
Convolution
mask
Using this convolution mask
on this convolution kernel
the final value of the pixel (4,2)
will be:
pixel (4,2) = 1/9(1,3) + 1/9(1,4)+ 1/9(1,45)
+1/9(2,3) +1/9(2,4) +1/9(2,5)
+1/9(3,3) +1/9(3,4) +1/9(3,5) X

44. 1/9 X
Convolution
mask
Using this convolution mask
on this convolution kernel
the final value of the pixel (5,2)
will be:
pixel (5,2) = 1/9(1,4) + 1/9(1,5)+ 1/9(1,6)
+1/9(2,4) +1/9(2,5) +1/9(2,6)
+1/9(3,4) +1/9(3,5) +1/9(3,6) X

45. Homework: What would be the effect of this mask? 1/9 0/9 3/9 X Using a
different
Convolution
mask... X
Homework:
What would be the
effect of this mask?

46. Convolution Calculations
Refer to additional information for examples to be worked

47. Introduce Gimp Open Image in GIMP... Adjust levels

48. Questions?