JPEG has played an important role in the image compression field for the past two decades. Quantization tables in the JPEG scheme is a key factor that is responsible for compression/quality trade-off. Finding the optimal quantization table is an open research problem. Studies recommend the genetic algorithm to generate the optimal solution. Recent reports revealed optimal quantization table generation based on a classical genetic algorithm (CGA). Although the CGA produces better results, it shows inefficiency in terms of convergence speed and productivity of feasible solutions. This paper proposes a knowledge-based genetic algorithm (KBGA), which combines the image characteristics and knowledge about image compressibility with CGA operators such as initialization, selection, crossover, and mutation for searching for the optimal quantization table. The experimental results show that the optimal quantization table generated using the proposed KBGA outperforms the default JPEG quantization table in terms of mean square error (MSE) and peak signal-to-noise ratio (PSNR) for target bits per pixel. The KBGA was also tested on a variety of images in three different bits values per pixel to show its strength. The proposed KBGA produces an average PSNR gain of 3.3% and average MSE gain of 20.6% over the default JPEG quantization table. The performance measures such as average unfitness value, likelihood of evolution leap, and likelihood of optimality are used to validate the efficacy of the proposed KBGA. The novelty of the KBGA lies in the number of generations used to attain an optimal solution as compared to the CGA. The validation results show that this proposed KBGA guarantees feasible solutions with better quality at faster convergence rates.
Abstract:The quantization table in the baseline Joint Photographic Experts Group (JPEG) algorithm plays an important role in compression/quality trade-off. Hence the detection of the optimal quantization table is viewed as an optimization problem. The genetic algorithm (GA) is an attractive optimization tool by many researchers for this application due to its ability in dealing with complex problems. In spite of its advantages, the GA requires more computation time to achieve an optimal solution if it has an expensive fitness evaluation. This paper proposes a problem approximation surrogate model (PASM) for fitness approximation to assist the GA in optimizing the quantization table for a target bits per pixel. This proposal reduces the computational time of the GA without any loss in performance. The PASM uses an image block clustering process and an indirect evaluation method to approximate the fitness value. The number of clusters in the clustering process may influence the performance of the PASM. A performance analysis with different number of clusters has been done and a suitable cluster number is identified with the help of measuring criteria such as mean squared difference, correct selection, potentially correct selection, and rank correlation. In addition, the results acquired from these measuring criteria are confirmed using statistical hypothesis tests such as Friedman's ANOVA and Wilcoxon signed rank. The PASM with suitable cluster number has been tested in a classical genetic algorithm and knowledge based genetic algorithm. Several benchmark images with different complexity levels have been examined in three different target bits per pixel to validate the performance of the PASM. The results proved that the PASM guarantees better results in terms of peak signal-to-noise ratio with a reduction in computational time.
Machine learning is one of the technologies coming to help the deployment of smart cities in all phases. The diagnosis is a crucial phase that comes to ensure the implementation of a project adapted to the reality of the city diagnosed; this step requires a significant financial commitment. This paper comes to deploy a frugal diagnostic approach of the smart environment component while using self-learning techniques. In addition, assessments are reported and regulatory maturity with respect to this new concept is explored through machine learning. In the near future machine, learning will play a crucial role in the implementation of this kind of concept.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.