Integration of Image Decomposition Methods and CNN for Image Classification

• DOI: 10.15388/lmitt.2025.6
• Publication Date: 2025-05-09T06:24:13Z
• Authors: Mahammad Ismayilov; Dalia Čalnerytė

This study explores integrating Haar wavelet decomposition techniques with convolutional neural networks for image classification on the MNIST dataset. The research demonstrates that without losing significant accuracy by applying 1-level, 2-level, and 3-level techniques, model can reduce dimensionality number of parameters required network model. During training, 1-level CNN results achieved optimal performance, demonstrating competitive computational efficiency compared to baseline approach highlights potential enhance performance limited resources.