Databases Reference
In-Depth Information
F I GU R E 3 . 16
Test images.
T A B L E 3 . 29
Compression using Huffman codes on
pixel values.
Image Name
Bits/Pixel
Total Size (bytes)
Compression Ratio
Sena
7.01
57,504
1.14
Sensin
7.49
61,430
1.07
Earth
4.94
40,534
1.62
Omaha
7.12
58,374
1.12
certain applications where it is necessary to know in advance how many bytes will be needed
to represent a particular data set.
The results in Table
3.29
are somewhat disappointing because we get a reduction of only
about
1
2
to 1 bit/pixel after compression. For some applications, this reduction is acceptable.
For example, if we were storing hundreds of thousands of images in an archive, a reduction of
1 bit/pixel saves many gigabytes in disk space. However, we can do better. Recall that when
we first talked about compression, we said that the first step for any compression algorithm
was to model the data so as to make use of the structure in the data. In this case, we have made
absolutely no use of the structure in the data.
