Image watermarking is a crucial technique used to protect digital media from unauthorized access and ensure copyright security. In cryptography, image watermarking is the process of encoding visible or hidden data onto digital photographs for uses like secure communication, copyright protection, and authentication. By identifying unlawful changes, it protects image integrity and aids in ownership verification. However, the effectiveness of a watermarking system depends significantly on its robustness against various attacks and understanding how different attacks can affect the quality of watermarked images. In this paper, the performance of a discrete cosine transforms (DCT) based image watermarking scheme is analyzed in the presence of different security attacks like salt-pepper noise, median and motion attacks. The performance is evaluated both visually and by MSE parameters.