Radar Development Kit

The Radar SDK, proprietary software for evaluating Infineon's XENSIV™ radar sensors, provides versatile accessibility through C/C++, Python, and Matlab. It encompasses a wide range of both basic and advanced radar algorithms.

Key components of the Radar SDK include:

• Radar SDK Library: This C/C++ library offers functions for configuring radar sensors and retrieving raw data. It also incorporates algorithms for Presence Sensing, Segmentation & Seamless Tracking, and Motion-Angle.
• Example Applications in C: These applications demonstrate the usage of the SDK and provide command-line tools for running algorithms on radar sensors.
• Python Wrapper and Example Applications: This component allows you to configure radar sensors, retrieve raw ADC data, and explore basic radar algorithms using Python.
• Matlab Wrapper and Example Applications: Similar to the Python wrapper, it enables you to configure radar sensors, retrieve raw ADC data, and explore basic radar algorithms using Matlab.
• Documentation: This resource provides usage information, building instructions, JSON configuration schema details, and guidance for sensor configuration.

Built using C++ and Qt, the Radar Fusion GUI is a cross-platform Graphical User Interface (GUI) compatible with Windows, macOS, and Linux. It functions as a versatile interface for Infineon's XENSIV™ radar demo boards, providing users with effortless visualization of both raw and processed data. Its core capabilities include streamlined radar device setup, real-time display of raw data and algorithm outputs, smooth data recording and playback, and the flexibility to export device configurations in multiple formats (such as JSON, and more). This tool is thoughtfully designed to enhance data visualization and empower users with an intuitive tool for graphical data analysis.

The Infineon Data Acquisition (ifxdaq) tool is designed to simplify the recording of high-quality sensor data sets, with a current focus on XENSIV™ 60GHz FMCW radar sensors. It supports simultaneous recording of sensor and camera data, with camera data generating labels for sensor data as the ground truth automatically. The labeled data can be used to validate the performance of radar-based applications or to train AI/ML algorithms.

To record ground truth, the user needs an Intel RealSense D455 camera. If a camera is not available, only radar data can be recorded. The recorded data is saved in a folder, where the radar data is stored in .npy format, a standard binary file format in NumPy. Radar configurations and metadata are stored in JSON format.