Please fill in your name

Mobile phone format error

Please enter the telephone

Please enter your company name

Please enter your company email

Please enter the data requirement

Successful submission! Thank you for your support.

Format error, Please fill in again


The data requirement cannot be less than 5 words and cannot be pure numbers

Multi-Sensor Fusion Perception in Autonomous Driving

From:Nexdata Date:2024-04-02

The era of Software-Defined Vehicles has arrived. Chips, operating systems, algorithms, and data together form the intelligent driving vehicles’ computing ecosystem. Algorithms and data are important tools to achieve advanced autonomous driving technology.

With the development of intelligent driving technology and the successful commercial application of intelligent vehicles, the gap in algorithms will shrink gradually. The role of data will continue to grow and become an important factor affecting intelligent driving technology.

The development of multi-sensor fusion perception requires the collection of various data such as images, videos, and 3D point clouds through multiple sensors, and finally achieves accurate recognition of objects near the vehicle through professional annotation.

In the field of automatic driving, Nexdata provides multi-sensor data collection services such as in-vehicle, roadside, and APA data, and supports 2D, 3D, 2D-3D fusion data labeling through self-developed data labeling platform. Our data solution can help intelligent vehicles quickly complete the whole process of perception-information processing-decision-making.

We provide flexible data collection solutions consist of hardware and software. For hardware, we can use single or multiple mechanical LiDARs. In addition, we use five or more high resolution industrial cameras, inertial navigation module of GPS/IMU fusion, and some other auxiliary devices.

The reason for using multiple LiDARs is that the point cloud data collected by multiple LiDARs is denser, and the data features are more abundant. Mutiple LiDars have different launch angles, which can reach the blind zone in a certain area. Five or more industrial cameras can achieve high frame rate and high resolution image acquisition with no blind zone.