The BYU Feature Matching dataset was created from real world images. This dataset is available publicly. There are 20 sets of images in a 60 MB zip file.   Each set consists of four image sequences. Each sequuence includes six images that have gone through one of four image transformations that incldue "blurring", "compression", "light variation", or "viewpoint change". The first of six images in each sequence is the original image and the subsequent images have increasing level of image transformation. Please double click on the image below to download.   We would appreciate you citing the source of this dataset.