This paper presents a novel hybrid image compression framework that integrates Discrete Wavelet Transform (DWT), a Wavelet-GAN encoder, vector quantization, and entropy coding to achieve high compression ratios with minimal loss in visual quality. The proposed method begins by applying DWT to the input image, extracting multi-resolution frequency sub-bands. A generative adversarial network (GAN) based encoder is then employed to capture the most significant features from the wavelet coefficients, effectively reducing redundancy. These encoded features undergo wavelet and vector quantization to further reduce data dimensionality. The quantized data is then encoded using entropy coding techniques, such as Huffman encoding, to generate a compact bitstream suitable for efficient storage and transmission. For image reconstruction, the process is reversed using entropy decoding, dequantization, and inverse DWT (IDWT) to obtain the final decompressed image. Experimental evaluations demonstrate that the proposed method provides an excellent trade-off between compression efficiency and image quality, making it particularly effective for applications in medical imaging, remote sensing, and real-time image transmission systems. We show that this work performs outstandingly with a Mean Squared Error (MSE) of 4.35, Peak Signal-to-Noise Ratio (PSNR) of 41.56 dB, Structural Similarity Index Measure (SSIM) of 0.9997, Compression Ratio (CR) of 18.92 and Bitrate of 0.47, indicating that it is both efficient and effective in maintaining image quality and achieving compression.