This is a set of functions for octave. I developed those functions for my personal research, so some of them may be kind of ad-hoc. Anyway, you might find some of them useful.
At the time of writing these lines, Imolib ships with next functions:
- imolib_block_dct
- Performs the (integer) DCT and inverse DCT transform for a matrix M in NxN blocks. M dimensions must be multiple of N. If N is 8, this function performs reasonably fast. Otherwise, it would probably be too slow for you.
octave:2> imolib_block_dct([1,2,4,5;3,4,5,6;4,5,6,7;6,7,8,9],2) ans = 5 -1 10 -1 -2 -0 -1 0 11 -1 15 -1 -2 -0 -2 0 octave:3> imolib_block_dct(ans, 2, true), ans = 1.00000 2.00000 4.00000 5.00000 3.00000 4.00000 5.00000 6.00000 4.00000 5.00000 6.00000 7.00000 6.00000 7.00000 8.00000 9.00000 - imolib_block_quantise
- Block-wise direct or inverse matrix quantisation. A quant matrix may be used as in MPEG-2 compression.
octave:9> imolib_block_quantise([1,2,4,5;3,4,5,6;4,5,6,7;6,7,8,9], 2, 1, false, [1,2;3,4]) ans = 1 1 4 3 1 1 2 2 4 3 6 4 2 2 3 2 octave:10> imolib_block_quantise(ans, 2, 1, true, [1,2;3,4]) ans = 1 2 4 6 3 4 6 8 4 6 6 8 6 8 9 8
- imolib_dct
- Extremely slow integer DCT for square matrices.
- imolib_fdct
- Optimised implementation for DCT and inverse DCT transforms for 8×8 matrices.
- imolib_idct
- Inverse of
imolib_dct. - imolib_load_pgm
- Loads pgm images.
- imolib_load_ppm
- Loads ppm images as three matrices (red, green, and blue).
- imolib_mpeg2_chroma_quant
- Returns the quantisation matrix for chroma blocks in MPEG-2 compression.
- imolib_quantise
- Direct or inverse matrix quantisation. Additionally to a global quantisation factor, you can use a per-coefficient quantisation matrix.
- imolib_upsample
- Naïve image upsampling.
octave:22> imolib_upsample([1 2 3; 4 5 6], 2) ans = 1 1 2 2 3 3 1 1 2 2 3 3 4 4 5 5 6 6 4 4 5 5 6 6
- imolib_yuv2rgb
- YUV to RGB (and vice versa) conversion.
You can download it from my github, or clone the repository git clone git://github.com/samuelrivas/imolib.git
