Skip to content

Latest commit

 

History

History
98 lines (62 loc) · 3.23 KB

README.md

File metadata and controls

98 lines (62 loc) · 3.23 KB

Capture Viewer

capture-viewer is a Luxonis Python tool designed to view and analyze data captured by OAK cameras. It provides functionalities to capture, display, and inspect stereo depth captures along with other types of camera data. It also allows the use of a REPLAY to regenerate data with different settings.

Requirements

  • Python (Tested on Python 3.11)
  • DepthAI version 2.29.0 or higher
  • Compatible with RVC2 OAK cameras (Tested on OAK-D and OAK-D-POE, compatible with Wide and PRO models)
  • To use the REPLAY feature, you need to have a camera connected to your PC

Ensure the required dependencies are installed by running:

pip install -r requirements.txt

How to Run

1. Capturing Data

See GUIDE

2. Viewing Data - Running the Capture Viewer

To start the capture viewer and view the captured data, run:

python capture_viewer.py -p DATA

Or specify your own capture folder:

python capture_viewer.py -p /path/to/DATA

Where DATA is the folder containing all the capture sessions. It will direct you to a session selection, where you can visualize individual sessions.

Optionally, you can add --ip to specify the IP of the device you want to run the replay on.

python capture_viewer.py -p /path/to/DATA --ip device_ip

REPLAY Feature

The REPLAY feature allows you to regenerate data using different settings or stereo algorithms. To use REPLAY, you need the left, right, and optionally color images from the camera. You can adjust various parameters like disparity algorithms and filter settings.

The REPLAY script can be found in the depth module and the user GUI is available in capture_viewer.py.

REPLAY features can only be used if you have an OAK camera connected to your PC via USB, or if you specify the IP of a network device.

REPLAY Output

The output from using REPLAY is automatically saved in path/to/DATA/capture_folder/replay_outputs
(e.g., /home/user/projects/capture-viewer/DATA/OAK-D-PRO_20241217125608/replay_outputs/20241217133208).

Each folder is named based on the time of generation and contains the following files:

  • config.json - Configuration used for depth generation.
  • depth_timestamp.npy and depth_timestamp.png - Depth generated with the given config from left_timestamp.npy and right_timestamp.npy.
  • pcl_timestamp.ply - Pointcloud generated from the depth. If the depth is aligned to the RGB camera, the pointcloud uses colors from the RGB camera; otherwise, it uses the JET colormap.

Adjusting window size

If the App window is too large/small for your screen, you can adjust window size. This will be improved in future versions of the app.

For one capture display: coak_capture_show.py

# change window size here
canvas_width = 1620
canvas_height = 880

For REPLAY: capture_viewer_tools/ReplayVisualizer.py

class ReplayVisualizer:
    def __init__(self, root, view_info, current_view):
        self.window = tk.Toplevel(root)
        self.window.title("Depth Visualization")
        self.window.geometry("2200x1200")  # adjust here

        max_image_width = 640  # adjust here
        max_image_height = 400  # adjust here
...