Method, apparatus, and system for removing outliers from road lane marking data

An apparatus comprising:

• at least one processor;
• and
• at least one memory including computer program code for one or more programs,
• the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following,
• process spatial data points representing road lanes to remove one or more miss-identified road lanes therefrom;
• and
• generate a road lane map layer based on the processed spatial data points;
• and
• provide the road lane map layer as an output,
• wherein the spatial data points are processed into a plurality of clusters, one or more clusters are filtered from the plurality of clusters to generate an initial filtered set of the spatial data points, and the filtering is based on determining that the one or more clusters is away from at least one main cluster by at least a threshold,
• wherein a degree of relationship is calculated among the spatial data points of the initial filtered set based on a distance metric to generate one or more group nodes of the initial filtered set, and wherein a group node is a center of a spatial data point group, regardless of density in the spatial data point group,
• wherein for each group node of the one or more group nodes, a probability that a spatial data point belongs in said each group node is determined based on a relative distance from the spatial data point to a center of said each group node, and filtering the spatial data point from said each group node based on determining that the probability is less than a threshold probability to generate a final filtered set of the spatial data points, and
• wherein the one or more miss-identified road lanes include the one or more clusters away from the at least one main cluster, the spatial data point with the probability less than the threshold probability, or a combination thereof.

A non-transitory computer-readable storage medium carrying one or more sequences of one or more instructions which, when executed by one or more processors, cause an apparatus to perform:

• monitoring one or more miss-identified road lanes determined from spatial data points overtime, wherein the spatial data points are received from one or more sources;
• computing a false rate based on the one or more miss-identified road lanes associated with the one or more sources;
• and
• providing the false rate as an output,
• wherein the spatial data points are processed into a plurality of clusters, one or more clusters are filtered from the plurality of clusters to generate an initial filtered set of the spatial data points, and the filtering is based on determining that the one or more clusters is away from at least one main cluster by at least a threshold,
• wherein a degree of relationship is calculated among the spatial data points of the initial filtered set based on a distance metric to generate one or more group nodes of the initial filtered set, and wherein a group node is a center of a spatial data point group, regardless of density in the spatial data point group,
• wherein for each group node of the one or more group nodes, a probability that a spatial data point belongs in said each group node is determined based on a relative distance from the spatial data point to a center of said each group node, and filtering the spatial data point from said each group node based on determining that the probability is less than a threshold probability to generate a final filtered set of the spatial data points, and
• wherein the one or more miss-identified road lanes include the one or more clusters away from the at least one main cluster, the spatial data point with the probability less than the threshold probability, or a combination thereof.