Basalt VIO

The example creates a pipeline to capture 640x400 stereo camera frames at 60 FPS and IMU data at 200 Hz, processes them with BasaltVIO for visual odometry, and streams the resulting transformations and images to a RerunNode for real-time visualization.

Setup

This example requires the DepthAI v3 API, see installation instructions.

Pipeline

React Flow

Source code

Python

C++

Python

GitHub

1import signal
2import time
3import depthai as dai
4from rerun_node import RerunNode
5# Create pipeline
6
7with dai.Pipeline() as p:
8    fps = 60
9    width = 640
10    height = 400
11    # Define sources and outputs
12    left = p.create(dai.node.Camera).build(dai.CameraBoardSocket.CAM_B, sensorFps=fps)
13    right = p.create(dai.node.Camera).build(dai.CameraBoardSocket.CAM_C, sensorFps=fps)
14    imu = p.create(dai.node.IMU)
15    odom = p.create(dai.node.BasaltVIO)
16
17    rerunViewer = RerunNode()
18    imu.enableIMUSensor([dai.IMUSensor.ACCELEROMETER_RAW, dai.IMUSensor.GYROSCOPE_RAW], 200)
19    imu.setBatchReportThreshold(1)
20    imu.setMaxBatchReports(10)
21
22    # Linking
23    left.requestOutput((width, height)).link(odom.left)
24    right.requestOutput((width, height)).link(odom.right)
25    imu.out.link(odom.imu)
26    odom.passthrough.link(rerunViewer.inputImg)
27    odom.transform.link(rerunViewer.inputTrans)
28    p.start()
29    while p.isRunning():
30        time.sleep(0.01)

Need assistance?

Head over to Discussion Forum for technical support or any other questions you might have.

ON THIS PAGE

Basalt VIO

Setup

Pipeline

Source code

Python

Need assistance?