Algorithm 3 APF method implementation procedure.

Input: Scene map $maze$, random start position $(x_0,y_0)$, random destination position $(x^d,y^d)$, positions set ${\Lambda }_{positions}$.

Output: A path between $(x_0,y_0)$ and $(x^d,y^d)$.

1: initialization: Set current position $(x,y)=(x_0,y_0)$, gray processing and binarization for the map $maze$.

2: Confirm the location of obstacles and calculate potential energy of each position.

3: while $|(x^d,y^d)-(x,y)|>ϵ$ do

4:  Record current position into ${\Lambda }_{positions}$.

5:  Find the position $(x,y)$ with the smallest potential energy among the eight positions around $(x,y)$.

6:  Perform $(x,y)=(x_{t+1},y_{t+1})$.

7: end while

8: return A path with all positions in ${\Lambda }_{positions}$.