In the modern transportation system, the number of automobiles in use continues to rise, and the safety of pedestrians on the roads has increasingly become a focus of attention. According to statistics, there are tens of thousands of casualties caused by vehicle-pedestrian collisions every year. How to effectively prevent such tragedies has become an important topic in the development of automotive safety technology. As a key component of the intelligent vehicle sensing system, the radar module is playing an irreplaceable role in the field of pedestrian detection.
Radar, that is, "Radio Detection And Ranging" (RADAR), detects information such as the distance, speed, and angle of the target object by emitting radio waves and receiving the reflected waves. The radar modules applied in vehicle pedestrian detection mainly operate in the millimeter-wave frequency band, commonly 24GHz and 77GHz. Millimeter-wave radars have many advantages. Their shorter wavelength allows the radar antennas to be made smaller, making it easier to integrate them into vehicles. At the same time, they are less affected by the weather. Whether it is rainy and foggy weather or a dusty environment, they can work relatively stably, which is beyond the reach of other sensors such as cameras.
When the millimeter-wave radar is working, the transmitter generates a high-frequency oscillation signal, and the millimeter waves are radiated outward through the antenna. After encountering a target object such as a pedestrian, part of the millimeter waves will be reflected back and received by the radar antenna. The receiver amplifies, mixes, and processes the received weak echo signal, and converts it into a low-frequency signal that is easy to analyze. By analyzing the frequency difference (i.e., the Doppler shift) between the echo signal and the transmitted signal, the radar can calculate the speed of the pedestrian. By using the time difference between the transmitted and received signals and combining it with the propagation speed of millimeter waves in the air, it can accurately determine the distance between the pedestrian and the vehicle. In addition, through an array composed of multiple antennas and using a specific algorithm, the angle at which the pedestrian is located can also be determined.
In actual vehicle pedestrian detection scenarios, the radar module faces a variety of complex challenges. In urban streets, the traffic flow is dense, and pedestrians shuttle among them. There are also a large number of stationary or slowly moving objects such as parked vehicles on the roadside and traffic signs. This requires the radar to be able to accurately distinguish pedestrians from other targets and avoid false positives. For example, at a crossroads, when pedestrians and turning vehicles appear simultaneously, the radar needs to instantly identify the motion states of different targets and provide accurate information to the vehicle control system so that it can make timely braking or evasion decisions.
To address these challenges, automotive radar technology is constantly evolving. On the one hand, at the hardware level, the resolution and detection accuracy of the radar are improved. For example, more advanced chip manufacturing processes are adopted to improve the radar signal processing capability; the number of antennas is increased and the antenna layout is optimized to enhance the angular resolution, thus enabling more accurate positioning of pedestrians. On the other hand, the software algorithms are continuously optimized. Machine learning and deep learning algorithms are introduced, allowing the radar to learn the characteristic patterns of pedestrians in different scenarios, such as the micro-Doppler characteristics when pedestrians are walking and the radar echo characteristics in different postures, so as to improve the accuracy and reliability of pedestrian recognition.
Take the two-chip cascaded imaging radar solution of Gatorland based on the Andes SoC chip as an example. This solution adopts a unique architecture of a quad-core CPU + digital signal processor (DSP) + dedicated radar signal processor (RSP), supports the innovative Flex-Cascading? patented technology, and realizes the imaging radar system through the cascading of two chips. Its maximum detection range reaches 320 meters, and it has excellent distance resolution and angular resolution capabilities. In the detection of vulnerable road users (including pedestrians), the excellent radio frequency performance of the 22nm CMOS process technology is used to meet the detection requirements of small targets; the DDM and coherent CFAR algorithms are adopted to obtain higher processing gain; the multi-band chirp technology is used to improve the distance and speed resolution to deal with slow targets; the dynamic windowing (SVA) technology is used to improve the influence of strong targets on the shielding of weak targets; the high-performance and low-time-consuming angular super-resolution algorithm solves the pain point of angular resolution; combined with micro-Doppler and AI technologies, it improves the pedestrian recognition ability, and has a significant effect on pedestrian detection in complex traffic environments.
The application of the radar module in vehicle pedestrian detection has greatly improved the active safety performance of automobiles. When the radar detects a pedestrian in front and there is a risk of collision with the vehicle, it will immediately send a signal to the vehicle control system to trigger the automatic emergency braking system (AEB), causing the vehicle to decelerate or stop within a short time to avoid the collision. Or through the warning system, such as the in-vehicle buzzer, dashboard warning lights, etc., it reminds the driver to pay attention to pedestrians and take timely measures. Many automobile manufacturers have already regarded the radar-assisted pedestrian detection function as an important part of the vehicle safety configuration. With the maturity of the technology and the reduction of costs, this function is gradually spreading from high-end models to more mid-range and low-end models.
However, the radar module is not perfect. In some extreme cases, such as encountering strong reflective objects like large metal billboards, the radar may be interfered with and false positives may occur. In a multipath propagation environment, the millimeter-wave signal may be received by the radar only after multiple reflections, resulting in deviations in the measurement of the target's position and speed. Therefore, to further improve the reliability of pedestrian detection, it is common to use the radar in combination with other sensors such as cameras. The camera can provide rich visual information and has a strong ability to recognize the appearance and texture features of objects, complementing the advantages of the radar in distance and speed measurement. Through the data fusion algorithm, by comprehensively processing the information from the radar and the camera, the accuracy and robustness of pedestrian detection can be effectively improved.
Looking to the future, with the continuous progress of science and technology, the radar module will usher in greater development in the field of vehicle pedestrian detection. On the one hand, the performance of the radar will be further improved, with a longer detection range, higher resolution, and stronger anti-interference ability, and it will be able to adapt to more complex traffic environments. On the other hand, the integration with other emerging technologies such as vehicle-to-everything (V2X) will be closer. V2X enables information interaction between vehicles, between vehicles and pedestrians, and between vehicles and infrastructure. The radar module can obtain more information about the surrounding environment and combine it with its own detection data to achieve more comprehensive and accurate pedestrian detection and early warning, and contribute more to the construction of a safe and intelligent transportation environment. As the invisible protector for vehicle pedestrian detection, the radar module is constantly evolving, safeguarding the safety of every pedestrian on the road and becoming an important driving force in the development of modern automotive safety technology.
