Tesla applies for new patent to improve Autopilot system’s obstacle recognition capabilities

June 19, according tomedia reports, earlier this month,media reported that a Tesla Model 3 that turned on Autopilot’s self-driving assisted driving system failed to identify a white truck that overturned in the middle of the road and crashed directly into the road, although no one was injured, but this indicates that the Tesla Autopilot system is still lacking in ability to identify obstacles. According to the latest reports, Tesla is also trying to improve autopilot’s ability to recognize obstacles with its automatic driver-assisted driving system, which it has obtained.

Tesla applies for new patent to improve Autopilot system's obstacle recognition capabilities

According tomedia reports, Tesla’s newly acquired patent, called “View-based Obstacle Enhancement Detection for Self-Driving Cars,” describes Tesla’s increased focus on obstacles that are critical to self-driving vehicles while reducing the collection of secondary image data.

Specifically, Tesla’s new patent focuses on information about obstacles in the direction of the vehicle and improves the ability to identify obstacles in the main direction.

As mentioned in the patent description, cameras, image sensors, sensors in a particular direction, such as image sensors located in front of the vehicle, and images of the true position of the vehicle in the direction of travel may be given, which may tend to depict information that must be paid attention to in vehicle navigation such as pedestrians, vehicles, obstacles, and so on, which needs to be focused on.

Tesla’s electric car, through multiple cameras or sensors, to identify and perceive objects in the real environment, cameras obtain edited images, some pedestrians, others are vehicles, animals or other obstacles, accurate identification of specific types of obstacles, the safety of people in the car is very important, for others, the system accurately and timely identification of obstacles is critical.