Straight-Line Intersection Point Detection Based on Curve Fitting of Hough Parameter Space

Straight-line intersection point detection plays an important role in machine vision and video analysis. For the actual problem including multi-line point, implicit point, local feature point rapid detection, the paper gives a Straight-line intersection detection method based on the Hough parameter space curve fitting. This method use the duality principle of Hough line detection, least squares sine curve fitting of the extreme points in Hough space, Reverse straight-line intersection point in X-Y coordinate system. The paper gives specific method of least squares sine curve fitting and extreme points choosing. Simulation result shows that the method has good and stable performance.