/pykinect-recorder

pykinect-recorder: RGB-D, IR record, playback, and inference for computer vision techniques

Primary LanguagePythonMIT LicenseMIT

pykinect-recorder



Description

The pykinect-recorder is an educational/industrial library that provides sensor recording (including audio), playback, and computer vision soultions through a python wrapper of the Azure Kinect Sensor SDK.

Recording and playback example below.

Documentation

You can find the API documentation on our website: https://pykinect-recorder.readthedocs.io/en/latest/index.html.

For details about API for Azure Kinect SDK please see Azure Kinect Sensor SDK github: https://github.com/microsoft/Azure-Kinect-Sensor-SDK.

Features

  • See RGB, IR, Depth, IMU and Audio data when recoding.
  • Control recording option (FPS, brightness, ...).
  • Change layout with drag and drop.
  • Playback recorded video.
  • 3D reconstruction viewer with streaming/recorded video.
  • Sync devices
  • screen zoom in-out
  • imu, microphone panel redesign
  • Recording audio.
  • Deep learning inference (mediapipe and native) with streaming/recorded video.
  • Intel RealSense
  • Zenmuse SDK for Python

Prerequisites

Environment

  • Windows 10 (Recommended)
  • Windows 11

Install Azure Kinect SDK

  • Make sure you download Azure Kinect SDK before using this repo.
  • SDK version '1.4.1' supported in release 0.9.3.
  • You can download Azure Kinect SDK here.

Installation

Using pip

python -m venv .venv
.venv/Scripts/activate.ps1
pip install pykinect-recorder
pykinect

Using .exe