论文部分内容阅读
文中针对空间曲线平滑这个图像理解与模式识别中的基本问题,在分析现有主要曲线平滑算法(如DouglassPeucker方法和固定窗口平滑算法)不足之处的基础上,利用信号平滑滤波的基本概念,基于曲线的参数方程和方向链码的原理,提出了一种基于Freeman链码和以弧长为参数的离散空间曲线的自适应平滑算法。该算法首先将空间曲线沿弧长方向进行分解;然后在分解后曲线的各个部分中,利用Freeman链码计算其平滑窗口的大小,并在窗口中进行平滑;最后,将各个部分合成,从而得到被平滑后的曲线。该算法解决了图像处理中曲线提取后存在的不光滑问题。最后给出了一些对比实验结果,证明了本算法的有效性
In this paper, the basic problems of image smoothing and pattern recognition in spatial curve smoothing are analyzed. Based on the analysis of the deficiencies of the existing main curve smoothing algorithms (such as DouglassPeucker method and fixed window smoothing algorithm), the basic concepts of signal smoothing filtering are used. Curve parameter equation and directional chain code principle, an adaptive smoothing algorithm based on Freeman chain code and discrete curve with arc length as parameter is proposed. Firstly, the space curve is decomposed along the arc length. Then the smoothing window is calculated by Freeman chain code in each part of the decomposed curve and is smoothed in the window. Finally, the parts are synthesized to obtain the The smoothed curve. The algorithm solves the problem of unsmooth curve extraction after image processing. Finally, some comparative experimental results are given to prove the effectiveness of this algorithm