Today's cars have gradually developed in terms of automatic assisted driving, coupled with the advent of the era of electrification, it is highly valued by major car companies.
In the field of automatic driving, sensors are indispensable settings, when your hands are free, then the sensor will become your eyes, for you to perceive all the objects around you, and the sensors used in automatic driving perception technology mainly include cameras, lidar and millimeter-wave radar.
And now in the choice of sensors, car companies are divided into two major gangs, one is Tesla led by the camera as the main hardware, to the visual-led automatic assisted driving development, and the other to lidar as the main hardware, on behalf of the car company is the domestic Xiaopeng, Xiaopeng is the world's first lidar to put on the car company, so in the end which of these two is better, in the end who is the best choice for automatic driving? It's not known, but it's been recently revealed that Toyota Motor's Woven Planet has joined Tesla in an attempt to advance autonomous driving technology without the use of expensive sensors such as lidar.
So what is the difference between lidar and camera?
First of all, let's talk about lidar, its advantage is that the detection distance is long and the accuracy is high, and the current perception ability of lidar is slightly stronger than the camera, and the accuracy is higher than the camera at night or in some extreme weather; its disadvantage is that the cost is high and it is difficult to mass-produce.
From the sensor point of view, the camera itself has an absolute advantage, the camera is smaller, easier to integrate into the design of the car, and the cost of the camera is lower, indirectly reducing the cost of automatic driving, but its disadvantage is that the perception in the dark environment will be limited, the accuracy and safety have decreased; in addition, in extremely bad weather, or other cases that cause the camera to not work, you need to rely on powerful algorithms to ensure the completion of perception processing.
Woven Planet's vice president of engineering said: "Cameras are being used to collect all kinds of information because the data collected from a small number of costly self-driving cars is not enough, and the accuracy of the data collected by low-cost cameras has now been improved to a level similar to that of high-end sensors." The cameras used by Toyota are now 90% cheaper than the sensors previously used and can be easily installed on different types of cars.
Although Toyota has now decided to take the pure visual route, it does not mean that it will completely abandon the multi-sensor fusion solution, considering the safety factors of users, toyota driverless taxis and other types of autonomous vehicles will still use laser radar and other sensors.
In fact, the perception fusion sensor represented by lidar, or the pure visual sensor represented by the camera, are the advanced exploration of automatic driving by car companies, and non-stop exploration can produce breakthroughs, Musk has said that this year Tesla can achieve full automatic driving through cameras.
Toyota has now stepped into the camp of Tesla, as a traditional car company, Toyota still has an advantage in many aspects. However, Toyota's experience in the field of automatic driving research should be relatively lacking, because Toyota set up an autonomous driving department in 2018, but has not disclosed the research progress in this area, so Toyota still has a long way to go after choosing a pure visual path.