Cloud segmentation is a vital task in applications that utilize satellite imagery. A common obstacle in using deeplearning-based methods for this task is the insufficient number of images with their annotated ground truths. This workpresents a content-aware unpaired image-to-image translation algorithm. It generates synthetic images with differentland cover types from original images, while preserving thelocations and the intensity values of the cloud pixels. Therefore, no manual annotation of ground truth in these images isrequired. The visual and numerical evaluations of the generated images by the proposed method prove that their qualityis better than that of competitive algorithms.