Map error detection system, map error detection method, and program

A map error detection system, comprising:

• a processor configured to execute instructions that
• calculate moving route information of a moving object by acquiring time-series positional information of the moving object and connecting pieces of the time-series positional information;
• compare the calculated moving route information with map information and to specify route information corresponding to the calculated moving route information among route information included in the map information;
• and
• perform clustering utilizing a number of positional information included within a predetermined range for an aggregation of the time-series positional information of the moving object in which the route information is unable to be specified when the route information is unable to be specified, determine a point at which a distribution density of the positional information becomes a maximum, and calculate the determined point as a map error candidate position which is a candidate for the route information not included in the map information,
• wherein the positional information in which the route information is unable to be specified includes
• the positional information included in the moving route information when the moving route information is determined to be different from the route information as a result of comparing the moving route information with the route information, and
• the positional information included in the moving route information when a moving direction indicated by the moving route information is different from a moving direction associated with the route information,
• wherein the processor is configured to execute instructions that calculate a plurality of clusters in the clustering such that each of the clusters includes a part of the aggregation of the positional information in which the route information is unable to be specified within the predetermined range and each of the clusters includes the part of the aggregation of the positional information that is different from each other, select the cluster including the largest number of the positional information among the calculated clusters, and determine a predetermined point in the selected cluster as the point at which the distribution density of the positional information becomes the maximum.

A map error detection method, comprising:

• a step of performing clustering utilizing a number of positional information included within a predetermined range for an aggregation of the positional information not corresponding to route information included in map information among time-series positional information of a moving object, determining a point at which a distribution density of the positional information becomes a maximum, and calculating the determined point as a map error candidate position which is a candidate for the route information not included in the map information,
• wherein the positional information not corresponding to the route information includes
• the positional information included in moving route information calculated by connecting pieces of the time-series positional information when the moving route information is determined to be different from the route information as a result of comparing the moving route information with the route information, and
• the positional information included in the moving route information when a moving direction indicated by the moving route information is different from a moving direction associated with the route information,
• wherein, in the step, calculating a plurality of clusters in the clustering such that each of the clusters includes a part of the aggregation of the positional information not corresponding to the route information within the predetermined range and each of the clusters includes the part of the aggregation of the positional information that is different from each other, selecting the cluster including the largest number of the positional information among the calculated clusters, and determining a predetermined point in the selected cluster as the point at which the distribution density of the positional information becomes the maximum are performed.

A non-transitory computer-readable medium that stores a program which causes a computer of a map error detection system to function as means for performing clustering utilizing a number of positional information included within a predetermined range for an aggregation of the positional information not corresponding to route information included in map information among time-series positional information of a moving object, determining a point at which a distribution density of the positional information becomes a maximum, and calculating the determined point as a map error candidate position which is a candidate for the route information not included in the map information,

• wherein the positional information not corresponding to the route information includes
• the positional information included in moving route information calculated by connecting pieces of the time-series positional information when the moving route information is determined to be different from the route information as a result of comparing the moving route information with the route information, and
• the positional information included in the moving route information when a moving direction indicated by the moving route information is different from a moving direction associated with the route information,
• wherein the means is configured to calculate a plurality of clusters in the clustering such that each of the clusters includes a part of the aggregation of the positional information not corresponding to the route information within the predetermined range and each of the clusters includes the part of the aggregation of the positional information that is different from each other, select the cluster including the largest number of the positional information among the calculated clusters, and determine a predetermined point in the selected cluster as the point at which the distribution density of the positional information becomes the maximum.