DCT compression,K-means compression,Wavelet compression
Evaluation of Compression Techniques
1. Starting with an image and a desired RMS error rate (these are variable selected by user).
2. Generic DCT compression:
1. Apply DCT on the entire image
2. Remove coefficients and find the RMS error
3. Repeat 2 until the desired RMS is met or exceeded
3. K-means compression:
1. Convert the color matrix MxNx3 matrix into MNx3
2. Apply k-means using variable k
3. Calculate the error for each k
4. Select k that exceeds or meets the desired k
4. Wavelet compression:
1. Apply wavelet decomposition using Haar wavelet at multiple levels. Decomposition will yield: A (scales), Vh, Vv, Vd (coefficients in horizontal, vertical, diagonal)
2. Reconstruct the images using one or more of the decomposition matrix
3. Compute the RMS error for each reconstruction.
4. Find the reconstruction that meets or exceeds the rms error
You may use MATLAB functions to generate DCT, kmeans, wavelet decomp and wavelet reconstruction, and haar wavelets.