chore: fix grids to grid

Signed-off-by: vividf <yihsiang.fang@tier4.jp>

calibrators/marker_radar_lidar_calibrator/README.md

@@ -25,7 +25,7 @@ After the background models for the lidar and radar are constructed, we compare
 All foreground radar objects are automatically categorized as potential reflector detections. For foreground lidar points, however, the [reflector](#radar-reflector) detection process involves more steps. We first apply a clustering algorithm on the lidar foreground points, then find the highest point in each cluster, and filter out the cluster if the highest point exceeds `reflector_max_height`. Next, we average all points within a `reflector_radius` from the highest point to estimate the center point of the reflector.
 
 The images below illustrate the process of radar background model construction and radar foreground extraction described in Step 1 and Step 2.
-The blue 3d voxels, which are shown in 2d grids in the images, are denoted as background voxels if any radar object falls into them during the background model construction. Once the background model is constructed, it becomes straightforward to extract the foreground objects in the calibration area. For the lidar, the background model and foreground extraction process are the same as the radar process described above.
+The blue 3d voxels, which are shown in a 2d grid in the images, are denoted as background voxels if any radar object falls into them during the background model construction. Once the background model is constructed, it becomes straightforward to extract the foreground objects in the calibration area. For the lidar, the background model and foreground extraction process are the same as the radar process described above.
 
 <table>
   <tr>