White Paper | Lidar 3D Point Cloud Processing and Sensor Fusion based on Intel® Architecture for C-V2X

JHC Product News

With the rapid development of technologies such as artificial intelligence, edge computing and mobile networking, the functions and performance of the C-V2X are constantly being improved, and it will play an important role in the future ITS.

Lidar 3D Point Cloud Processing and Sensor Fusion based on Intel® Architecture for C-V2X

Recently, JHCTECH has collaborated with Intel and Leishen to issue a White Paper based on Roadside Sensing and Roadside Edge Computing solutions for the C-V2X industry, introducing the JHCTech® Roadside MEC Equipment based on the 11th-Generation Intel® Core™ Processors and Intel® Distribution of OpenVINO™ Toolkit, which is used to support the 3D point cloud processing based on deep learning and the Sensor Fusion for Leishen® All-in-One Roadside Sensing Equipment (Lidar and Camera).

JHCTech® Roadside MEC Equipment

Emphasized in the white paper, in the applications of C-V2X, the Multi-access Edge Computing (MEC) plays an extremely important role.

JHCTECH has developed a new KMDA-3301 Roadside MEC Equipment, which is powered by 11th-Gen Intel® Core™ Processors. Based on Intel® Architecture, the KMDA-3301 can provide powerful and reliable general-purpose and AI computing power for various use cases of C-V2X. This enables us to perform real-time and efficient analysis of information from different types of traffic sensors and fuse the results, which significantly improves the safety and efficiency of the ITS. The C-V2X applications for traffic safety have strict requirements for end-to-end (E2E) latency. The Roadside MEC has a better guarantee for reducing the E2E latency.

JHCTECH‘s MEC equipment solution for C-V2X has been deployed in more than 10 cities such as Beijing, Shanghai, Guangzhou, Shenzhen, Chongqing, Changzhou, etc. The features of high performance and reliable high computing power have made JHCTECH’s MEC equipment an ideal solution for the harsh roadside environments in C-V2X application.

Roadside Edge Computing: Promoting the Sensor Fusion in C-V2X applications

In the area of Roadside Sensing, the commonly used traffic sensors include cameras, lidars and mmWave radars. The technical solutions introduced in this paper involve the first two.

Sensor Fusion: Lidar and Camera

The Sensor Fusion can be achieved through the "division of labor" between the camera and the lidar, where the camera is used for the object classification and the lidar is used for the object detection.

The Roadside MEC Equipment based on Intel® Architecture can perform the inference based on deep learning or analysis based on traditional computer vision on the 2D video images and 3D point clouds, respectively, and merge the results of the two by Sensor Fusion algorithms.