Title : Using discrete cosine transform to encode approximation wavelet subband. +CD

number: 
1883
English
department: 
Degree: 
Imprint: 
Physics
Author: 
Sara Adnan Mahmood
Supervisor: 
Dr.Loay E. George
year: 
2008
Abstract:

Various compression methods have been proposed to achieve high compression ratios and high image qualities in low computation time relatively. In this research work a combined transform coding scheme was proposed, the adopted system utilize both Discrete Cosine Transform and Wavelet transform. The advantages of both transforms were taken into consideration to encode the image. First, the system transform the color components of the image from (RGB) to (YUV), the U and V bands are downsampled due to their poor spatial resolution, and then the wavelet transform is applied on each color band separately. Some spatial coding steps are applied on detail coefficients (like hierarchal uniform quantization, Run Length Encoding (RLE), shift coding) to gain more compression. The approximate coefficients are coded by using (DCT), uniform quantization, shift coding. Also, some analysis tests were done to study the performance of the established system, and the effects of the involved coding parameters on the system were investigated. The test results indicated that the proposed scheme give high compression ratio and good fidelity measures (MSE and PSNR). Moreover, the proposed scheme was found need 0.7 second to compress color images of size (256x256) without making significant degradation in image quality.