A method is proposed for extracting attitude data from streaked star images using traditional image processing methods. The process enables collection of both attitude and angular velocity estimates from streaking star images, where traditional star identification methods would produce poor results. Star streak endpoints are localized as "corner-like" portions of a streak. A vector cross-product based method is developed to produce a proper grouping in time for endpoints. Star identification and single point attitude determination methods can be applied to each set of endpoints, retrieving two sets of data from a single image. Monte Carlo results are presented, and the implications of the results are discussed. Multiple corner detection methods are considered and compared. Future work needed to mature the process is discussed. Results indicate an endpoint detection accuracy of less than a tenth of a pixel for a camera angle to the rotation axis of greater than 20 degrees and streak lengths up to 40% of the field of view.