Lane boundary line recognition device and computer-readable storage medium storing program of recognizing lane boundary lines on roadway

A computer-implemented method, performed by a processor, for recognizing a lane boundary line comprising:

• detecting lane boundary line candidates of a roadway on the basis of frame images of the roadway around an own vehicle transmitted from an in-vehicle camera;
• judging whether the detected lane boundary line candidate corresponds to a lane boundary line of a branch road;
• and
• calculating feature values comprising a curvature of the detected lane boundary line candidate, and recognizing the lane boundary line on the basis of the calculated feature values, wherein
• recognizing the lane boundary line further comprises:
• removing the lane boundary line candidate, which has been judged to correspond to the lane boundary line of the branch road, from the lane boundary line candidates;
• calculating a curvature of the lane boundary line candidate on the basis of an estimated rate of change of the curvature of the lane boundary line candidate;
• resetting the calculated curvature to a past curvature of the lane boundary line candidate calculated a predetermined number of frame images prior when it is detected that the lane boundary line candidate corresponds to the lane boundary line of the branch road;
• and
• resetting the estimated rate of change of the curvature of the lane boundary line candidate to zero, wherein
• the curvature of the lane boundary line candidate is calculated on the basis of an estimated rate of a second differential of the curvature of the lane boundary line candidate, and
• the estimated value of the second differential of the curvature of the lane boundary line candidate is reset to zero when it is judged that the lane boundary line candidate corresponds to the lane boundary line of the branch road.

A system for recognizing a lane boundary line comprising:

• a central processor;
• a computer-readable storage medium;
• and
• a set of computer-executable instructions stored on the computer-readable storage medium that cause the central processor to:
• detect lane boundary line candidates of a roadway on the basis of frame images of the roadway around an own vehicle transmitted from an in-vehicle camera;
• judge whether the detected lane boundary line candidate corresponds to a lane boundary line of a branch road;
• and
• calculate feature values comprising a curvature of the detected lane boundary line candidate, and recognizing the lane boundary line on the basis of the calculated feature values, wherein recognizing the lane boundary line further comprises:
• removing the lane boundary line candidate, which has been judged to correspond to the lane boundary line of the branch road, from the lane boundary line candidates;
• calculating a curvature of the lane boundary line candidate on the basis of an estimated rate of change of the curvature of the lane boundary line candidate;
• resetting the calculated curvature to a past curvature of the lane boundary line candidate calculated a predetermined number of frame images prior when it is detected that the lane boundary line candidate corresponds to the lane boundary line of the branch road;
• and
• resetting the estimated rate of change of the curvature of the lane boundary line candidate is reset to zero, wherein
• the curvature of the lane boundary line candidate is calculated on the basis of an estimated rate of a second differential of the curvature of the lane boundary line candidate, and
• the estimated value of the second differential of the curvature of the lane boundary line candidate is reset to zero when the branch judgment section judges that the lane boundary line candidate corresponds to the lane boundary line of the branch road.