Notice: Trying to access array offset on value of type null in /home/gadget/public_html/wp-content/plugins/really-simple-facebook-twitter-share-buttons/really-simple-facebook-twitter-share-buttons.php on line 318
It would seem that it won’t be long now until we see a driverless car pass by, telling us that we don’t know how to drive. Thanks to several ongoing projects, we will hear more about the robot cars in the nearby future. Cambridge wiz techs have developed new gizmos for autonomous vehicles, thus ensuring that such a car can see better.
Apart from being a very good punchline from an adaptation of Red Riding Hood, the autonomous vehicles now actually have bigger eyes in order to distinguish better between several traffic instances.
According to professor Robert Cipolla, the head of the research project, the new systems will not only ensure that the car doesn’t mistake you for a runway, but they will also be capable of cutting back the costs of making such a high-tech car.
The new models of driverless vehicles will include two new programs. The first one is called SegNet, and, according to the team, it’s far more accurate than any laser sensor or image interpreter. SegNet is capable of scanning any image depicting a road and identify the certain instance found in a picture.
Furthermore, through the new system computer system, the car will be fit to distinguish between several environmental entities such as the main road, pedestrians, street signs and other vehicles in the area. Even though the system is still being tested out, it has shown a tremendous potential in terms of navigation.
The professor and his team have enlisted the help of many undergrad students in order to improve SegNet. And so, for several days, a couple of students fed many pictures to the program. All and all, the program has been able to distinguish between different pixels with a 90 percent accuracy.
Cipolla declared that no image analysis system was able to obtain such a high score. The second innovation in the field of driverless automobiles it the location system. According to project’s abstract, thanks to the new locative algorithm, the car will now be able to figure out its relative position in the world, without using GPS devices.
This system holds a tremendous potential taking into consideration the fact that are many area with low, limited or no GPS signal, such as tunnels or country roads.
The Cambridge professor is very confident that the new project will step up the robotization process. Thus, somewhere in the nearby future we are bound to see more driverless vehicles on the road.