Accurately identifying resident space objects (RSOs) within optical space imagery poses a key challenge in the realm of Space Situational Awareness (SSA). For unresolved imagery, the primary issue is distinguishing RSOs from stars and other objects (or aberrations) that may be present. In this project we explore a first-principles approach to identifying RSOs based on the nature and degree of streaking in an image. Different objects in an image will streak to different extents, depending on each object’s motion relative to the observer’s platform. It is this knowledge, combined with appropriate processing of the image, that allows effective discrimination between RSOs and stars. The proposed technique is developed to operate on a single image as opposed to a multi-frame collect.