Hybrid DCT-DWT with Adaptive Quantization for High-Efficiency Image Compression

Khac Trung Kien

Abstract: This study introduces a hybrid image compression approach that integrates two-level Discrete Wavelet Transform (DWT) with block-based Discrete Cosine Transform (DCT), combined with adaptive quantization. The proposed framework applies DCT to approximation coefficients derived from Haar wavelet decomposition and scales quantization matrices based on target quality parameters. Evaluated using standard grayscale images, the method demonstrates compression ratios ranging from 85.89:1 to 255:1 with consistent visual fidelity, as indicated by PSNR values between 22.24 and 27.65 dB and SSIM scores from 0.75 to 0.84. When compared to JPEG and JPEG2000, the approach achieves notably higher compression at lower bit rates. Its applicability spans bandwidth-limited domains such as satellite imaging and mobile transmission, where aggressive compression with preserved visual quality is essential.

Keywords: Image compression, Hybrid transform, Adaptive quantization, DCT, DWT

How to Cite?: Khac Trung Kien, "Hybrid DCT-DWT with Adaptive Quantization for High-Efficiency Image Compression", Volume 14 Issue 12, December 2025, International Journal of Science and Research (IJSR), Pages: 268-271, https://www.ijsr.net/getabstract.php?paperid=SR251129091423, DOI: https://dx.doi.org/10.21275/SR251129091423