Combining Haar transform and row-column directional RLE in image compression

Haar transform, the simplest wavelet transformation, converts the data into wavelet coefficients. The output of the transformation generates larger numbers of Zeros. It acts as a basic step for data compression. The small values are further quantized to Zero for more appearance of Zeros in the data. Run-length encoding is a simple method of compression for repeating characters and is suitable here for repetitive Zeros. The quantized data are encoded in both horizontal and vertical direction run-length encoding. In this paper, we propose an algorithm that combines the Haar transform with the row-column directional run-length coding for a high rate of compression. The analysis of the results shows that the approximation of Haar coefficients increases the number of Zeros with a small increase in MSE in the transformed data. The visual or perceptual quality does not change so much.