Object detection

1. A system, comprising:

a Lidar sensor; and

a processor programmed to:

generate a two-dimensional grid map including a plurality of grid cells based on data received from the Lidar sensor;

detect, in the two-dimensional grid map, a low-density area comprising a plurality of grid cells with an intensity less than a threshold;

determine dimensions of the low-density area;

determine that the detected low-density area is unexpected upon failing to identify a second object that casts a shadow matching at least one of a location, the dimensions, and a shape of the unexpected low-density area;

based on the detected low-density area and the determined dimensions of the low-density area, determine that the unexpected low-density area represents a physical object;

select an expanded area of the two-dimensional grid map including the low-density area;

apply a filter with a second threshold to the expanded area of the two-dimensional grid map;

upon identifying an occupied area within the expanded area, determine that the occupied area represents the physical object;

detect a second object based on at least one of Lidar sensor data and map data, and

upon determining that the detected low-density area is a shadow cast by the second object, ignore the detected low-density area.

2. The system of claim 1 , wherein the processor is further programmed to:

detect a second area, within a specified distance from the low-density area, comprising a second plurality of grid cells with an intensity exceeding a second threshold; and

based on the detected low-density area and the second area, determine that the second area represents a physical object.

3. The system of claim 1 , wherein the processor is further programmed to:

determine a plurality of volumes, wherein each volume has a bottom placed on a ground surface and a top spaced upwardly from a ground surface; wherein the bottom matches a grid cell of the two-dimensional grid map; and

determine an intensity of each grid cell, matching a bottom of a volume, based on a number of Lidar reflections received from points within the respective volume.

4. The system of claim 1 , wherein determining that the low-density area represents the physical object further comprises using map data.

5. The system of claim 4 , wherein the map data include at least one of (i) map data received from a remote computer, and (ii) historic map data received from a vehicle object detection sensor.

6. A method, comprising:

generating a two-dimensional grid map including a plurality of grid cells based on data received from a Lidar sensor;

detecting, in the two-dimensional grid map, a low-density area comprising a plurality of grid cells with an intensity less than a threshold;

determining dimensions of the low-density area;

determining that the detected low-density area is unexpected upon failing to identify a second object that casts a shadow matching at least one of a location, the dimensions, and a shape of the unexpected low-density area;

based on the detected low-density area and the determined dimensions of the low-density area, determining that the unexpected low-density area represents a physical object;

selecting an expanded area of the two-dimensional grid map including the low-density area;

applying a filter with a second threshold to the expanded area of the two-dimensional grid map;

upon identifying an occupied area within the expanded area, determining that the occupied area represents the physical object;

detecting a second object based on at least one of Lidar sensor data and map data, and

upon determining that the detected low-density area is a shadow cast by the second object, ignoring the detected low-density area.

7. The method of claim 6 , further comprising:

detecting a second area, within a specified distance from the low-density area, comprising a second plurality of grid cells with an intensity exceeding a second threshold; and

based on the detected low-density area and the second area, determining that the second area represents a physical object.

8. The method of claim 6 , further comprising:

determining a plurality of volumes, wherein each volume has a bottom placed on a ground surface and a top spaced upwardly from a ground surface; wherein the bottom matches a grid cell of the two-dimensional grid map; and

determining an intensity of each grid cell, matching a bottom of a volume, based on a number of Lidar reflections received from points within the respective volume.

9. The method of claim 6 , wherein determining that the low-density area represents the physical object further comprises using map data.

10. The method of claim 9 , wherein the map data include at least one of (i) map data received from a remote computer, and (ii) historic map data received from a vehicle object detection sensor.