In this paper, we present an inventive way of reducing salt-and-pepper noise in color images employing deep learning-based denoising convolutional neural network (DCNN). This kind of noise, containing random pixels of white and black color, is detrimental to image quality and poses challenges to image processing tasks. The technique here takes advantage of the strong ability of DCNNs to extract features effectively for the purpose of discerning noise while maintaining key aspects within the picture. Specifically, the network design is such that it can handle the intricacies involved with color image denoising by using specifically tailored layers and loss functions. There are many benefits to the method proposed when compared with usual filtering techniques. To start with, a deep learning model can be trained to detect and filter out complex noise patterns that cannot be detected by humanly designed filters. Another advantage is that DCNNs are able to work directly on color images, without losing any color data during denoising processes. The paper presents some experiments on a benchmark dataset illustrating the efficiency of the suggested DCNN algorithm and its contrast to other modern denoising methods. These findings imply that DCNN is better in terms of noise reduction and quality preservation for colored images that have been damaged by salt-and-pepper noise, hence offering a more promising solution.