Digital Images Retrieval Using Iterative Technique

Use of efficient image restoration method gains most interest due to its close relation to the improvements happen in the fields of compression and communications. Image restoration tries to improve the quality of degraded image by undoing the blurring and reducing the noising using the de-convolution technique for producing an estimate image is approaches the original one. The adopted image retrieval technique uses the Meads filter that working in the frequency domain for achieving the recovered image. In order to test the performance of the adopted method, a gray image is used in the implemented test with a pixel resolution of 8bit.The practical work includes two phases: degradation phase and retrieval phase. The degradation contains two stages within, they are: the blurring and noising. While the retrieval phase depends on processing the amount of blurring that defined by the blurring parameter (σ), and then rejecting an amount of the noise that determined by the noise parameter (SNR). Different values of blurring and noising parameters are estimated when applying the Mead's technique for achieving the retrieved image. Results showed high matching between the retrieved image and the original one. Quality measures prove the retrieved image converge the original image with little differences in between, where the amount of similarity score is 92%.