针对传统 Canny 算子在平滑图像、梯度计算和阈值选取方面存在缺陷，论文提出了一种改进方案。该方法使用带权中值滤波代替高斯滤波，使用3×3邻域四方向偏导数代替2×2邻域二方向来求取梯度幅值和方向，使用梯度直方图确定高、低阈值代替人为随意设定。仿真实验结果表明，改进的 Canny 算子提高了检测精度，连续性更好。%In view of the defects of the traditional Canny operator in smoothing image ,gradient calculation and thresh-old selection ,an improved scheme is proposed in this paper .The method uses a weighted median filtering instead of the Gaussian filter ,3 × 3 neighborhood four direction partial derivative is used to replace the 2 × 2 neighborhood two direction to obtain the gradient magnitude and direction ,and uses Gradient histogram to determine high and low threshold instead of man-made arbitrarily set .The simulation experimental results show that the improved Canny operator to improve the detec-tion accuracy and have better continuity .
展开▼
