To efficiently extract the textual information from color degraded document images is an important research topic. Long-term imperfect preservation of ancient documents has led to various types of degradation such as page staining, paper yellowing, and ink bleeding; these degradations badly impact the image processing for information extraction. In this paper, we present CCDWT-GAN, a generative adversarial network (GAN) that utilizes the discrete wavelet transform (DWT) on RGB (red, green, blue) channel splited images. The proposed method comprises three stages: image preprocessing, image enhancement, and image binarization. This work conducts comparative experiments in the image preprocessing stage to determine the optimal selection of DWT with normalization. Additionally, we perform an ablation study on the results of the image enhancement stage and the image binarization stage to validate their positive effect on the model performance. This work compares the performance of the proposed method with other state-of-the-art (SOTA) methods on DIBCO and H-DIBCO ((Handwritten) Document Image Binarization Competition) datasets. The experimental results demonstrate that CCDWT-GAN achieves a top two performance on multiple benchmark datasets, and outperforms other SOTA methods.