As we all know, in the field of autonomous driving, there has been a dispute over technical routes before.
Tesla takes the pure visual route, that is, only uses cameras to collect information, and then uses algorithms to achieve automatic driving, Tesla believes that since people can drive with their eyes, then they must also use cameras, if not, it means that the technology is not good, and it can be improved again.
Many companies believe that the camera only obtains plane information, cannot judge the distance, etc., and will be affected by the weather, darkness, etc., LiDAR can solve this problem.
The representative enterprise of this solution is Huawei, which uses lidar, and as for the high cost of lidar? Huawei said that the price can be reduced by expanding the scale of use and upgrading technology.
These two sets of plans have been fighting before, and Musk is a fool to use lidar, which is too expensive. Cameras can also improve ISP technology, collect information in dark conditions, and also obtain distance information.
The lidar solution party complained about Tesla, all kinds of misjudgments, and it was clear that the pure vision was backward, and it was not admitted.
In fact, to be honest, there is no right or wrong plan, let alone advantages and disadvantages, one is to look at your own technical level and business model. Pure vision requires more powerful algorithms, and lidar needs more powerful hardware, stacking hardware is easy, but stacking algorithms is relatively difficult.
At the same time, the final point of contention is actually whether the ISP technology of the camera can also achieve the effect of lidar faster, or whether the price of lidar is more expensive to reduce to the camera.
If the camera achieves the effect of lidar and the price is still cheap, who will use lidar?
Similarly, if the price of lidar drops to the camera, who doesn't use lidar?
In fact, cost is also one of the key items to consider, if the cost can not be supported, then you know that lidar is better, and you can only use the camera solution.
For example, recently, Huawei also learned from Tesla and came up with a purely visual autonomous driving technology.
This set of technology is put on the newly released Zhijie S7 low-configuration version, called Huawei ADS2.0 basic version. In this scheme, there is no lidar, and the visual scheme is used.
This set of solutions can also realize the intelligent navigation of national high-speed and urban fast, and can also achieve intelligent parking, such as off-line parking spaces, oblique parking spaces and custom parking spaces.
It can be seen that going around and around, in the end, Tesla is more powerful. It has long realized that in the end, the development of autonomous driving technology must be purely visual, with low cost and high solution advantages. As for why you can't do it, it's because you can't do it, and when you can do it, you can do it.
Therefore, we can boldly predict that when the technology develops to a certain extent, the pure visual intelligent driving solution will inevitably replace the lidar solution and become the optimal solution for autonomous driving.