<p>The star sensor is recognized as the most precise attitude determination sensor for space applications. This sensor establishes attitude by capturing images of stars within its field of view. The calculation of attitude via the star sensor involves several algorithms, among which star identification is the most critical. Incorrect star identification can result in erroneous attitude determinations. One of the lesser-explored star identification algorithms employs singular value decomposition (SVD) of a matrix formed by the directional vectors of a star cluster. This paper provides a survey of various star identification techniques based on this concept, along with simulations, analyses, and enhancements of these methods. Additionally, the impact of varying star cluster sizes and different field of view dimensions on the algorithm's performance has been examined across multiple methodologies. The identification process is framed as a two-step procedure utilizing singular values and singular vectors. Furthermore, the prevalence of duplicate sets within the databases of different methods has been investigated. In the concluding section of this paper, novel approaches for eliminating dimmer stars from images and selecting brightest stars based on the characteristics of singular value decomposition are proposed.</p>