Four Points that Provide a Solution in Railway Line Detection and Tracking

The novel use of the Euclidean distance between consecutive pixel intensity values is presented as a solution for railway line detection and tracking. The Euclidean distance between consecutive pixel intensity values, a searching strategy and a mathematical tool, in the form of collinearity testing, are the key elements of this method of railway line detection and tracking. The Euclidean distance between consecutive pixels’ intensity values defines edges. Four points are marked where significant changes in pixel intensity values occur on two horizontal lines across an image; two on each line. These points are used for detecting and tracking railway line tracks using the collinearity test of the cross product of two vectors. The four points are also used as the control points for correcting perspective distortion in an image, better known as transforming the image into a bird’s-eye view using the Top View Transformation model. Finally, the four points are used in calculating the changes in the pitch angle as the altitude of the railway line track bed rises and falls. The changes in the pitch angle are used in a model to stabilize the transformation of the image into the bird’s-eye view. The searching strategy, which is the cornerstone of this algorithm refers to the placement of the two lines across the image. An equation is presented that governs how far apart the two lines should be based on the curvature of the railway line track.