Disclosed herein are a method, system, and computer program product for creating a scene model of a scene, for separating a background and a foreground of at least one frame in an image sequence depicting the scene. The method includes the following steps, dependent upon a stability measure associated with the scene model being less than a stability threshold: (i) capturing a frame in the image sequence (305); (ii) recording the captured frame to a stored video sequence (335), dependent upon a recording session being active (330); (iii) updating an initial scene model, based on the captured frame (340); (iv) determining the stability measure associated with the scene model (340); and (v) providing a feedback, dependent upon the stability measure being less than a stability threshold (370). Then, dependent upon the recording session being active, the method captures at least one subsequent frame in the image sequence and records each captured subsequent frame to the stored video sequence (335). Finally, the method produces a scene model (380) that distinguishes background from foreground in the scene, based on the updated initial scene model.