This　paper　proposes　an　instantaneous　rotational　speed　(IRS)　sensing　method　that　utilizes　a　circumferential　constant-density-sine　fringe　pattern　(CCDSFP)　captured　by　a　linear-array　sensor　(LAS).　A　novel　mathematical　imaging　model　has　been　established　to　develop　a　mapping　relationship　between　the　correlation　coefficient　(CC)　value　of　adjacent　frame　fringe　signals　and　the　IRS.　The　influence　of　various　factors　on　the　sensing　performance　was　thoroughly　investigated　through　simulations　and　experiments.　The　results　indicate　that　this　method　demonstrates　satisfactory　measurement　accuracy　within　the　measurement　range　from　0　rpm　to　the　maximum　measurable　rotation　speed.　The　innovation　of　this　method　lies　in　using　only　a　single　line　of　fringe　signal　captured　by　LAS　for　sensing　the　rotation　speed,　which　offers　significant　advantages　in　rotation　speed　sensing　and　computational　efficiency　compared　to　area-array　camera-based　method.　Moreover,　the　new　speed　encoding　and　estimation　methods　exhibit　very　high　sensitivity　and　accuracy　for　very　low　and　high　rotation　speed.　Compared　to　encoders　based　on　pulse　counting　during　a　unit　time　or　time　interval　estimation　between　two　pulses,　the　proposed　method　provides　higher　temporal　resolution　and　angular　resolution　for　rotation　speed　measurement.　Furthermore,　this　novel　sensing　method　offers　a　non-intrusive　approach　that　has　no　additional　mass　effect　on　the　shaft,　and　it　could　be　used　for　the　development　of　new　kind　of　IRS　sensor.　©　2024　IEEE.