Staircase Detection Algorithm Based on Projection-Histogram

Detecting a staircase is the first step in achieving autonomous ambulation of unmanned vehicles and helping the blind to climb the stairs. In this paper, based on the detection of the edge of the stairs, a method of using the statistical properties of projection histogram to identify the stairs is proposed. To do this, first use a directional Gabor filter and a Canny edge detector on the staircase image to eliminate the effects of lighting and detect the edges of the stairs. Remove noise in the image of the edge of the stairs by performing a filter operation. Then extract the decreasing length of the horizontal edge as the final image. Finally, histogram statistics are performed after the pixel values of the adjacent lines of the image are superimposed, and the rule is given according to the law reflected in the histogram to determine whether the staircase is a stairway.