一种基于梯度极值的边缘检测算法

A DIFFERENTIAL EXTREMUM BASED EDGE DETECTION ALGORITHM

  • 摘要: 分析了经典微分边缘检测算法存在的问题,提出了一种基于局部区域微分极值的边界检测算法,该算法的特点是在局部区域边界检测窗口内直接提取边界象素,分割的边界为真实边界,边界连续性好,适合于照度不均匀图象的分割.实例给出了令人满意的结果.

     

    Abstract: The paper analyzes the existent question about the differential algorithm for image segmentation first, and then proposes a differential extremum based edge detection algorithm. The feature of the algorithm lies in that the edge pixel is directly extracted at a local window for the edge detection, the extracted edge is a true edge, and the edge possesses good continuity. The algorithm is suited for-segmentation of the image having inhomogeneous-illumination. The segmentation experiment gaves a satisfactory result.

     

/

返回文章
返回