Code Samples

Code samples are used for automated testing. They are also a great starting point for the DepthAI SDK, as different component functionalities are presented with code.

ColorCamera

• FFC Camera Visualization - Preview FFC Cameras

• Camera Control - Demonstrates RGB camera control from the host

• Camera Preview - Preview color, right, left and depth frames

• Camera Control with NN - Control camera (focus, exposure) with NN detections

• Mono Camera Preview - Preview mono cameras with manual 400p resolution

• Preview All Cameras - Preview all cameras connected to the OAK device

• RGB and Mono Preview - Preview RGB and mono cameras

• Camera Rotated Preview - Demonstrates how to rotate the camera previews

Mixed

• API Interoperability Example - Demonstrates interoperability between the DepthAI API and the SDK

• Car Tracking Example - Demonstrates how to run inference on a pre-saved video

• Collision Avoidance - Demonstrates how to run collision avoidance

• Speed Calculation Preview - Demonstrates how to calculate speed of detected objects in the frame`

• Switch Between Models - Demonstrates how to switch between models

• Sync Multiple Outputs - Demonstrates how to sync multiple outputs

IMU

• IMU Demonstration - Demonstrates how to use and display the IMU

• IMU Rerun Demonstration - Demonstrates how use and display the IMU in Rerun

NN

• Age-Gender Inference - Demonstrates age-gender inference

• Custom Decode Function - Demonstrates custom decoding function

• Emotion Recognition - Demonstrates emotion recognition

• Face Detection RGB - Run face detection on RGB camera

• Face Detection Mono - Run face detection on mono camera

• Human Pose Estimation - Run human pose estimation inference

• MobileNet Encoded - Pass encoded color stream to the NN (MobileNet)

• Neural Network Component - Run color camera stream through NN (YoloV7)

• Object Tracking - Tracking objects in the frame

• Roboflow Integration - Demonstrates how to use Roboflow platform to train a custom model

• Spatial Detection - Perform spatial detection with at MobileNet model

• YOLO SDK - Run YoloV3 inference on the color camera stream

Pointcloud

• Pointcloud Demo - Preview pointcloud with rerun viewer

Recording

• Encode Multiple Streams - Demonstrates how to encode multiple (color, left, right) streams and save them to a file

• Preview Encoder - Record color camera stream and save it as mjpeg

• MCAP Recording - Record color, left, right and depth streams and save them to a MCAP

• MCAP IMU Recording - Record IMU and depth streams and save them to a MCAP

• Hardcode Recording Duration - Record color camera stream for a specified duration

• ROSBAG Recording - Record IMU, left, right and depth streams and save them to a ROSBAG

• Stereo Recording - Records disparity stream

Replay

• Object counting on images - Count number of objects on a folder of images (cycle through images every 3 sec)

• People Tracker on Video Replay - Run people tracker on a pre-saved video

• Face Detection Inference on Downloaded Image - Run face detection on a downloaded image

• Vehicle Detection on a Youtube Video - Run vehicle detection on a Youtube video stream

• Looped Replay - Replay a pre-saved video in a loop

Stereo

• Stereo Preview - Display WLS filtered disparity map

• Auto IR Brightness - Demonstrates the use of auto IR brightness function

• Stereo Control - Demonstrates stereo control (median filter, decimation factor, confidence threshold) from the host

• Stereo Encoding - Demonstrates how to encode stereo stream and visualize it

Streaming

• ROS Publishing - Publish color, left, right and IMU streams to ROS

Trigger Action

• Custom Trigger Action - Demonstrates how to set a custom trigger action

• Custom Trigger - Demonstrates how to set a custom trigger

• Person Record - Demonstrates how to record a person when a person is detected

Visualizer

• Visualizer Demo - Demonstrates how to use the visualizer

• Visualizer Callback Function - Demonstrates how to set the visualizer callback function