Image Compression based on Fixed Predictor Multiresolution Thresholding of Linear Polynomial Nearlossless Techniques

Abstract

Image compression is a serious issue in computer storage and transmission, that simply makes efficient use of redundancy embedded within an image itself; in addition, it may exploit human vision or perception limitations to reduce the imperceivable information
 Polynomial coding is a modern image compression technique based on modelling concept to remove the spatial redundancy embedded within the image effectively that composed of two parts, the mathematical model and the residual.
 In this paper, two stages proposed technqies adopted, that starts by utilizing the lossy predictor model along with multiresolution base and thresholding techniques corresponding to first stage. Latter by incorporating the near lossless compression scheme of first stage that corresponding to second stage.
 The tested results shown are promising in both two stages, that implicilty enhanced the performance of traditional polynomial model in terms of compression ratio , and preresving image quality.

Compressing Face Images Using Genetic and Gray Wolf Meta-heuristic Algorithms Based on Variable Bit Allocation

“…Based on the CIE database, simulation results related to accuracy, average PSNR, and SSIM parameters are presented in Figures 11,12,and 13. As shown in Figure 14, the total amount of processing computing time for genetic and gray wolf meta-heuristic algorithms is calculated by dividing bit rate values from 0.5 to 1 into 16, 32, and 64 blocks. The meta-heuristic algorithm of genetics and gray wolf is used in this article to identify significant blocks and assign appropriate bits to each block.…”

“…The popularity of meta-heuristic methods can be attributed to their flexibility, derivative-free mechanism, and ability to avoid getting stuck in local optima. These methods [12] are relatively simple and are founded on very simple concepts. This article also discusses the gray wolf algorithm, which is based on collective intelligence and the life of gray wolves [13].…”

“…Figure 10, for example, shows the amount of total processing time for genetic and gray wolf meta-heuristic algorithms with 16, 32, and 64 block allocations for bit rates ranging from 0.2 to 1.Based on the CIE database, simulation results related to accuracy, average PSNR, and SSIM parameters are presented inFigures 11,12,and 13. As shown in Figure14, the total amount of processing computing time for genetic and gray wolf meta-heuristic algorithms is calculated by dividing bit rate values from 0.5 to 1 into 16, 32, and 64 blocks.…”

Image Compression using Polynomial Coding Techniques: A review

“…The use of hard/soft thresholding techniques of hierarchical scheme DWT Haar base to compress residual image of linear polynomial coding techniques adopted by Noor [26] (2015) to compress gray scale images then the same principle extended to compress color images by Ghadah et al [27] (2016). Where the former used scalar uniform quantizer for approximation sub band (LL) with values between 2 to10, along hard/soft thresholding for details sub bands (LH, HL, HH) with values between 20 to 40 and two standard grayscale square images (Lena and Woman) of sizes 256x256, the compression ratio of hard thresholding double the soft one but with less quality compared to soft that characterized by highly preserving quality.…”

“…Rasha [17] Mixed between linear and nonlinear polynomial coding techniques Utilized linear or non-linear models according to block nature where compression ratio is between 8 -13, with PSNR values exceeding 32 dB. 5 Noor [26] and Ghadah et al [27] Hard/soft thresholding utilized to compress details sub bands for gray and color images Residual image in gray or color base compressed using DWT hard/soft thresholding with efficient utilization of the frequency domain along with mixing thresholding techniques where the compression ratio for color images is between 5-12 with PSNR values between 28dB to 30dB. 6 Shymaa [28] Near lossless scheme adopted to compress residual efficiently Enhanced compression performance between 9 -11 with pleasing visual quality.…”