This paper introduces KinZ, an Azure Kinect toolkit for Python and Matlab. This toolkit encapsulates the Azure Kinect's main functionality in an easy-to-use class with high-level methods. This software is written mostly in C/C++ with a pybind interface for Python and Mex functions for Matlab. It provides access to color, depth, infrared cameras, a mapping between cameras, colored point clouds, IMU sensor data, body tracking with 32 body joints plus 32 hand keypoints. We showed that KinZ applications could run in real-time at 1080p resolution, with the benefit of a high-level language reducing the learning curve and allowing a significant decrease in development time for prototyping and research.

本文介绍了KinZ，这是一个适用于 Python 和 Matlab 的 Azure Kinect 工具包。该工具包将 Azure Kinect 的主要功能封装在一个具有高级方法的易于使用的类中。该软件主要用 C/C++ 编写，带有用于 Python 的 pybind 接口和用于 Matlab 的 Mex 函数。它提供对颜色、深度、红外摄像机、摄像机之间的映射、彩色点云、IMU 传感器数据、带有 32 个身体关节和 32 个手部关键点的身体跟踪的访问。我们展示了KinZ应用程序可以以 1080p 分辨率实时运行，高级语言的好处是减少了学习曲线，并显着减少了原型设计和研究的开发时间。