Device, system, and method for generating occupancy grid map

An occupancy grid map generation system comprising:

• a plurality of vehicles equipped with a detection device configured to detect an object therearound, a current location detection unit configured to detect data of a current location, and an occupancy grid map generation device for generating an occupancy grid map that indicates, for each of a plurality of cells of a grid into which a region around the current location is divided, an occupancy probability that an object occupies the cell;
• a probe data map server configured to acquire, from each of the plurality of vehicles, detection result data from the detection device and the data of the current location, wherein
• the occupancy grid map generation device of each of the plurality of vehicles comprises:
• a data acquisition unit configured to acquire, from the detection device, detection result data of the object;
• an occupancy probability updating unit configured to, based the detection result data, successively update the occupancy probability for each of the plurality of cells calculated at a previous time to calculate a latest occupancy probability;
• a map data acquisition unit configured to acquire probe data map around the current location of the vehicle from the probe data map server;
• and
• an updating weight calculation unit configured to acquire data about presence or absence of an object in each of the plurality of cells around the current location based on the probe data map, and determine a weight of the detection result data used by the occupancy probability updating unit to update the occupancy probability for the cell, based on a degree of match between the data about presence or absence of an object in the cell and the detection result data from the detection device, the updating weight calculation unit being configured to, when the degree of match between the detection result data and the data about presence or absence of an object based on the probe data map is high, calculate the weight of the detection result data to be greater than when the degree of match is low.

An occupancy grid map generation method for generating an occupancy grid map that indicates, for each of a plurality of cells of a grid into which a region is divided, an occupancy probability that an object occupies the cell, the occupancy grid map generation method comprising:

• acquiring, from a detection device that detects an object therearound, detection result data of the object;
• based the detection result data, successively updating the occupancy probability for each of the plurality of cells calculated at a previous time to calculate a latest occupancy probability;
• acquiring map data therearound;
• acquiring data about presence or absence of an object in each of the plurality of cells therearound based on the map data, and calculating a weight of the detection result data to update the occupancy probability for the cell, based on a degree of match between the data about presence or absence of an object in the cell and the detection result data;
• and
• based on the detection result data and the calculated weight, successively updating the occupancy probability for each of the plurality of cells calculated at the previous time to calculate a latest occupancy probability.

A non-transitory computer-readable storage media having instructions stored thereon that, when executed by a computer, cause the computer to perform an occupancy grid map generation method for generating an occupancy grid map that indicates, for each of a plurality of cells of a grid into which a region is divided, an occupancy probability that an object occupies the cell, the occupancy grid map generation method comprising:

• acquiring, from a detection device that detects an object therearound, detection result data of the object;
• based the detection result data, successively updating the occupancy probability for each of the plurality of cells calculated at a previous time to calculate a latest occupancy probability;
• acquiring map data therearound;
• acquiring data about presence or absence of an object in each of the plurality of cells therearound based on the map data, and calculating a weight of the detection result data to update the occupancy probability for the cell, based on a degree of match between the data about presence or absence of an object in the cell and the detection result data;
• and
• based on the detection result data and the calculated weight, successively updating the occupancy probability for each of the plurality of cells calculated at the previous time to calculate a latest occupancy probability.