Abstract
With the ever-growing volume of multimedia data, efficient image compression has become essential for storage, transmission, and real-time applications. Advanced signal processing techniques have significantly enhanced the performance of image compression models by exploiting spatial and frequency domain redundancies. This paper explores the evolution of image compression from classical transform coding to deep learning-based approaches integrated with signal processing paradigms such as Discrete Wavelet Transform (DWT), Principal Component Analysis (PCA), and learned convolutional autoencoders. Comparative analysis demonstrates how signal processing augments model accuracy, computational efficiency, and visual quality.

This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright (c) 2022 Dr. Elena V. Petrova (Author)