Uniformity in point features distribution of feature extraction algorithm at remote sensing image automated registration is a challenging problem. For the purpose of remote sensing image automated registration, a multi-feature image automated registration algorithm is presented based on Scale Invariant Feature Transform (SIFT), Harris-Laplace and Maximally Stable Extremal Region (MSER). Multi-feature and secondary matching greatly promotes the number of matching points, screening by distance makes matching points distribute uniformly and reasonably, and local mutual information takes matching points more accurate. Finally, a high quality (uniform space distribution, accuracy matching points) automated image registration is achieved.