Discovering subtle alterations of pairs of images taken from the same scene at different time intervals is called minute change detection problem. To achieve this goal, we have developed a framework that captures and highlights minute changes in digital images that are otherwise hidden to the human eye. Moreover, unnoticeable differences from image pairs that are taken at different time intervals with similar viewing conditions are detected. Although our framework's application areas cover a wide variety of different disciplines, from medicine to security, weather forecasting, urban planning, and monitoring natural disasters, in this study our focus was real-time minute change detection and tracking on biomedical and satellite images. Real-time performance in cases such as medicine is crucial; we enhance this approach by using the extensive computational power of the graphical processing unit (GPU). Our experimental results in detection of subtle changes in light microscopic images of mouse MC3T3-E1 osteoblastic cells grown in culture with the resolution of 2600x2060 and remote sensor images performed by the GPU computations illustrate that our algorithm detects infinitesimal differences on images in real-time.