Moving Object Classification Using 3D Point Cloud in Urban Traffic Environment

Table 2

Shape-based features.

Index	Description	Formula
f6	The sum of the area of the fitting rectangles of the horizontal projection points at each layer
f7	The average of the area of the fitting rectangles of the horizontal projection points at each layer
f8–f15	The density of the point cloud at each layer
f16	The area of the fitting rectangle of the horizontal projection points	W
f17-f18	The length and width of the fitting rectangle of the horizontal projection points	—
f19	The standard Cartesian dimension of the horizontal projection points
f20–f27	The mean of inscribed angles at each layer (the proximity to a circle)
f28–f35	The standard variance of inscribed angles at each layer
f36–f43	The circle fitting level at each layer (sum of squared residuals of the vertical distances between the points and the least square fitting circle)
f44–f51	The radius of the circle fitted by the horizontal projection of the points at each layer	Radius (circle x^2 + y^2+ Dx + Ey + F = 0)
f52–f59	The line fitting level at each layer (sum of squared residuals of the vertical distances between the points and the least square fitting line) [12]