Research On B-spline Fitting And Optimization Of Image Contour

The curve fitting problem has always been a key research topic in computer graphics and computer aided geometric design(CAGD).On the academic side,many scholars have done a lot of research on this.B-spline curve has excellent properties.Fitting discrete data points with B-spline is widely used in industrial production and still has important research significance.The fitting of the image contour is expressed in the form of a curve.Under a given error,the image contour is fitted with less curve segments and time.The result of curve fitting should retain the characteristic points of the image contour and ensure its rationality.This article first uses cubic quasi-uniform B-spline to fit the discrete data points of the image contour.The first and end points of discrete data points are selected as the first and last control vertices of the cubic quasi-uniform B-spline curve,and the least square algorithm is used to determine the distance between the internal control vertices of the cubic quasi-uniform B-spline curve and the first and last control vertices.The chord length parameterization assigns initial parameter values to each discrete data point.When the error is large,the point with the largest error is selected to divide the discrete data points into two subsets,and two cubic quasi-uniform B-splines are recursively fitted to the new one Sub-curve.The B-spline curve has excellent continuity and provides more degrees of freedom during segmental fitting.When the error is within a certain range,the Newton iteration method is used to obtain better chord-length parameterized parameter values,which reduces the control vertex.Numerical experiments show that it has certain advantages in calculation time and the number of application curve segments.Because the B-spline curve has excellent continuity,when the B-spline curve is used to fit an image contour with sharp-angled feature points,the fitted curve is easy to lose the sharp-angle feature of the image contour.Unreasonable phenomena such as self-intersection and curve distortion may occur in the process of applying B-spline curve to fit the contour of low-resolution images.This paper introduces three prior knowledges of length priority,avoiding self-intersection and avoiding the smoothing of sharp corners to restrict the B-spline curve fitting image contour.Numerical experiments show that the introduction of prior knowledge can solve the above-mentioned unreasonable phenomena in the process of fitting image contours,and obtain more reasonable and beautiful image contours.