Detect walls in the map created by gmapping

     When the robot is powered on , the original orientation is not certain . That means robot does not know how to walk along the wall.  The current way XiaoMi robot used to walk along the wall from the original position where the robot powers on is scanning the surroundings by rotating about 180 degree to find the closest point and walk towards to that point until the obstacle collision occurs. If we can get the angle between the wall and reference coordinate , then we can make the robot rotate to parallel of the wall which is ready to start the path plan. So I designed a scratch that using image processing methods on the static map would generate all lines represent all walls information including the length , the orientation , the end point position and formula of line.

     The algorithm is based on the HoughTransform method which is used to detect the lines on the binary value image. So the steps following stand for the whole processing .

Snapshot: