The Effect of Using Different Color Spaces On Haar Wavelet Transformed Images

Abstract

Wavelets provide a powerful set of tools for handling fundamental problems in science and engineering, such as image compression, edge detection and fingerprint compression and image de-noising. This paper applies the Haar wavelet on various types of images with different sizes after converting the colored image) RGB) to grayscale Image (LC color space which is based on luminance (L) and chrominance (C) values) like YIQ, YUV and YCbCr color spaces to see how it affects on the quality based on the PSNR value. This type of transform gives some degradation in quality, so this research helps the researchers to decide the best LC color space that have to be used when compressing images with Haar wavelet transform on various types of images before any manipulation to reduce the file size like Quantization, Shift Coding or other methods.The results of the conducted tests that applied on different types of images indicated that the difference between images in the YIQ and YUV color spaces is too small, so the difference between YIQ and YCbCr is taken into consideration in all comparisons. The tested results proved that the YCbCr color space is the best to use in images when they have sharpen edges or in highly detailed images and the difference in the PSNR is more than (1dB), this color space is also preferred in big images, while the effect of the three LC color spaces is not mentioned in the images with small sizes and less detailed images and any of these spaces can be used without affecting to quality of the image. The tests proved that the YIQ and YUV transform take less encoding time except in the detailed images.