You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+3-3Lines changed: 3 additions & 3 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -165,7 +165,7 @@ python controller.py --debug
165
165
166
166
### Mobile Base Accuracy
167
167
168
-
The marker detection setup should output 2D robot pose estimates with centimeter-level accuracy. For instance, our setup can reliably pick up small Lego Duplo blocks (32 mm x 32 mm) from the floor. Inaccurate marker detection can be due to many reasons, such as inaccurate camera alignment or suboptimal camera settings (see `get_video_cap` in [`utils.py`](server/utils.py)). Also note that the mobile base motors should be calibrated (`.motor_cal.txt`) for more accurate movement.
168
+
The marker detection setup should output 2D robot pose estimates with centimeter-level accuracy. For instance, our setup can reliably pick up small Lego Duplo blocks (32 mm x 32 mm) from the floor. Inaccurate marker detection can be due to many reasons, such as inaccurate camera alignment or suboptimal camera settings (e.g., exposure and gain, see `get_video_cap` in [`utils.py`](server/utils.py)). Also note that the mobile base motors should be calibrated (`.motor_cal.txt`) for more accurate movement.
169
169
170
170
### Arm Accuracy
171
171
@@ -186,8 +186,8 @@ To kill all processes using the occupied ports, you can use the [`clear-ports.sh
186
186
```
187
187
188
188
For reference, here are all of the ports used by this codebase:
4. Get the serial numbers of the two overhead cameras using either of the following commands:
39
+
40
+
```bash
41
+
lsusb -v -d 046d:0843 | grep iSerial
42
+
udevadm info -n video0 | grep ID_SERIAL_SHORT
43
+
```
44
+
45
+
5. Modify [`constants.py`](constants.py) to reflect your setup. These values may potentially need to be modified:
46
+
47
+
*``ROBOT_HOSTNAME_PREFIX``
48
+
*``CAMERA_SERIALS``
49
+
*``CAMERA_HEIGHT``
50
+
*``ROBOT_WIDTH``
51
+
*``ROBOT_HEIGHT``
52
+
*``ROBOT_DIAG``
53
+
38
54
## Usage
39
55
40
56
The following sections describe the process of setting up marker detection and object detection from scratch. To get started with a setup that already works, see the "TidyBot Quickstart" section of the main [`README`](../README.md).
41
57
42
58
We use two ceiling-mounted overhead cameras to detect fiducial markers installed on robots in the scene to get real-time 2D robot poses with centimeter-level precision. The camera we use is the [Logitech C930e webcam](https://www.amazon.com/Logitech-C930e-1080P-Video-Webcam/dp/B00CRJWW2G). Note that even though the C930e supports 1080p, we run it at 720p, as we found that 720p is faster and was sufficient to get centimeter-level accuracy.
43
59
44
-
The code uses the serial number of your camera as an identifier for`cv.VideoCapture` and for saving and loading camera parameters. On Ubuntu 20, you can get the serial number of your camera using either of the following commands:
45
-
46
-
```bash
47
-
lsusb -v -d 046d:0843 | grep iSerial
48
-
udevadm info -n video0 | grep ID_SERIAL_SHORT
49
-
```
60
+
The code uses the serial number of your camera as an identifier for`cv.VideoCapture` and for saving and loading camera parameters.
0 commit comments