METHOD FOR ALIGNING A PLURALITY OF SINGLE DRIVE MAPS
DRIVE
June 10, 2026
The invention discloses a Method for aligning a plurality of single drive maps (SDM1, SDM2), each single drive map (SDM1, SDM2) containing a plurality of data objects (6) associated with corresponding fixed objects (4) of a bordered traffic area (1), said data objects (6) being based on a data collection by a respective vehicle (mv1, mv2) when driving through said bordered traffic area (1), wherein for a first single drive map (SDM1), based on its respective data objects (6), a first set of first key points (pj) is determined, each key point being (pj, qj) indicative of a data object (6), by means of a key point determination algorithm (8), and for each first key point (pj), a feature descriptor (fd) indicative of the surroundings of the respective first key point (pj) is determined by means of a feature determination algorithm (10), wherein for a second single drive map (SDM2), based on its respective data objects (6), a second set of second key points (qj) is determined by means of said key point determination algorithm (8), and for each second key point (qj), a feature descriptor (fd) indicative of the surroundings of the respective second key point (qj) is determined by means of said feature determination algorithm (10), wherein for each of the first key points (pj), a corresponding second key point (qc(j)) is determined based on the respective feature descriptors (fd) of said first and second key points (pj, qj), wherein a rigid transform (R, t) for mapping the second single drive map (SDM2) is derived, said rigid transform (R, t) being determined on the basis of said correspondence (c(j)) of the first key points (pj) and the respective second key points (qc(j)).
Discussion in the ATmosphere