Three-dimensional (3D) feature-matching techniques, which are essential for progress towards an automated feature-based procedure, have attracted considerable research attention in both the photogrammetry and computer vision communities. This study introduces a novel matching approach, called RSTG, that comprises four major phases: rotation alignment; scale estimation; translation alignment; and geometry checks. These steps efficiently determine a feature-based correspondence and frame transformation between datasets. RSTG analyses the similarity and relative geometry of features by employing feature observations and their uncertainty; this allows different types of features to be matched exclusively or simultaneously. This study validates the proposed method with both simulated and real datasets, demonstrating its effectiveness with satisfactory matching rates in a diverse range of feature-based point cloud registration tasks. ). Retrieval precision, however, is reduced when object shapes contain abundant selfsimilarities. Studies have attempted to simplify matching problems by first eliminating rotation and translation discrepancies with principal component analysis and centres of gravity. These methods, however, work under extremely specific conditions and only obtain a coarse estimation of the transformation because of geometric and point-to-point constraints (Kim et al., 2011). At the computational level, least squares techniques that minimise spatial distances between homologous primitives are frequently implemented to acquire matching solutions. Consequently, a non-linear calculating procedure is required which calls for additional processes, resulting in a more complex operation (Xu and Li, 2000;Gruen and Akca, 2005). To the authors' knowledge, most existing feature-based methods concentrate only on pairing a specific kind of geometric primitive (Heuel and F€ orstner, 2001;Eden and Cooper, 2008;Kamgar-Parsi and Kamgar-Parsi, 2011).This study proposes a matching approach called RSTG (rotation; scale; translation; geometry) to retrieve corresponding 3D feature counterparts and frame transformations from unsorted datasets. Scenes in urban areas contain various objects, such as buildings, signs, cars, trees and poles, which are generally composed of basic primitive shapes. The feasible features in RSTG comprise points, straight lines and planes. Each feature can be used exclusively, or in combination, as long as the minimum number of features for the matching solution is satisfied. The versatility of feature usage offers high flexibility and facilitates dealing with scene complexity. Moreover, RSTG weights feature observations, allowing multiple features to be managed in an improved manner throughout the matching processes. A hierarchical matching strategy is also administered to increase the success rate and reliability of matching, and to reduce computational complexity.A valid procedure for acquiring features is a prerequisite to any feature-based technique. In this study, geometric features are automatica...