Hybrid DCT/Quantized Huffman compression for electroencephalography data

In this paper, We implement the Discrete Cosine Transform (DCT) coding (lossy compression method) followed by the proposed coding technique which called “Quantized Huffman Coding” in order to minimize the EEG data size. Therefore, adding a lossless compression algorithm after the lossy compression is a good idea to get a high compression ratio with acceptable distortion in the original signal. Here, we use DCT encoder followed by either quantized Huffman Coding or Run Length Encoding (RLE) then compare between them. Our work shows that, at the same Root Mean Square Error (RMSE), the quantized Huffman coding outperforms the RLE in some aspects as Compression Ratio (CR) and time consumed in compression and decompression, but Structural Similarity Index (SSIM) is the same for the two techniques.