Springbot: A Prototype Autonomous Vehicle and Its Algorithms for Lane Detection
a) This article is unique in that it is a discussion of current trends and a revised method in lane
detection and tracking. While this does not relate directly to the autonomous cruise control the
idea of lane detection and tracking is another part of intelligent vehicle systems and another
system we might need to consider interactions with.
b) One of the constraints for the proposed system is extracting and dealing with the curvature
of the road/earth. They use a series of 3D to 2D to 1D extraction methods to detect the lane at
further distances than previous techniques. Other limiting factors include running time of the
algorithm as well as detection of other vehicles on the road.
c) The articles is more of a comparison of different algorithms of lane detection and therefor is
lacking on the discussion of safety features. The algorithm itself does improve safety through its
ability to detect lanes at further distances than previous algorithms.