Abstract: During their acquisitions astronomical CCD images are affected by a variety of noises, among which cosmic rays sometimes directly affect useful information in the images. It is very important to effectively identify and remove the cosmic rays noise for the extraction of information from an astronomical image. For the three present algorithms to remove cosmic rays——the Laplacian edge detection algorithm,the fast algorithm based on histogram, and the universal noise removal Algorithm, we use them to process CCD images taken at the Yunnan Observatory with the 1-meter telescope with either actual or simulated cosmic-ray hits. We compare and analyze the levels of effectiveness and complexity of these algorithms. Processing results show that the Laplacian edge detection algorithm can most accurately detect cosmic rays in images of stars and galaxies. We also discuss how to accurately substitute the values of the pixels affected by cosmic rays.