In addition to considering the characteristics of hardware devices, compatibility with the bus interior environment must also be taken into account. After a comprehensive market survey of embedded computing modules, NVIDIA’s Jetson Xavier NX was selected as the core computing module. The choice of Jetson Xavier NX is based not only on its high performance and compact size but also on its ability to adapt to the bus interior environment. This processor features high-temperature resistance, vibration resistance, and electromagnetic interference protection, enabling stable operation in harsh onboard environments, with excellent computing performance and energy efficiency. This choice meets the performance requirements of the monitoring system while achieving efficient video processing in limited space. Furthermore, Jetson Xavier NX supports a rich set of interfaces and communication protocols, allowing seamless integration with existing hardware devices in the bus. Through these interfaces, data exchange and information sharing with other vehicle systems can be achieved, such as integration with the onboard navigation system, linking location information with video data, and providing more comprehensive monitoring and analysis capabilities. The NX module integrates a comprehensive deep learning inference computing framework, which is compact and powerful, and can be seamlessly integrated into a broader Jetson platform. During the system assembly process, in addition to the core computing unit, the selection and configuration of accessories such as video capture cards, carrier boards, and connection cables must also be considered. The reasonable pairing of these accessories can optimize system performance and ensure stable connections and data transmission between components. In summary, the selection of Jetson Xavier NX as the core computing module considers not only its high performance and compact size but also its compatibility and stability in the bus interior environment. By reasonably configuring accessories and interfaces, seamless integration with the onboard system can be achieved, enhancing the overall effectiveness of the monitoring system and providing reliable technical support for safe operations in the public transportation sector. The Jetson Xavier NX algorithm board and its key components are shown in the figure below.
The system uses the industrial-grade carrier RTSO-6002E, specifically designed for the Jetson Xavier NX module. This carrier board has stable performance and low power consumption, capable of reliable operation within a wide working range, and provides multiple external interfaces to ensure that computing results can be effectively transmitted and sent, thus meeting the demands of the bus environment. The model RTSV-6911i video capture card has also been selected to match the output signals of the cameras inside the bus. This video capture card has superior performance, featuring a Mini PCI Express interface, supporting 8 channels of NTSC/PAL video input and 8 channels of D1 full bitrate video capture. The RTSV-6911i also provides various functions: including directly outputting the captured video stream to system memory or local display, and converting recorded video data into various formats required by applications, such as system display, image analysis, and processing. The integration of these functions allows the system to more effectively process and analyze video data inside the bus, providing reliable technical support for safety monitoring.
This computing unit is designed and assembled with three main components: the carrier board, video capture card, and core computing motherboard, achieving the goals of video capture and automatic image processing. Each component plays a crucial role in the entire experimental setup, enabling effective data analysis and interpretation.
WeChat Official Account QR Code
WeChat Official Account: Artificial Intelligence Perception Information Processing Algorithm Research Institute
