基于Canny算子的白酒显微图像边缘提取

文章提出了一种改进的Canny算子和数学形态学相结合的白酒显微图像边缘检测方法。首先采用二维高斯函数的一阶偏导数同时构造x和y方向滤波器计算梯度幅值,然后通过四阈值边缘检测方法进行边缘定位,最后引入了数学形态学对边缘进行融合。实验仿真表明,该方法对酒分子结构边缘检测精度较准、抗噪性能良好,能有效的克服图像边缘模糊与断点,提高了边缘的连续性。

Spirits Microscopic Image Edge Detection Based on Canny Arithmetic Operators

An edge detection approach of spirits microscopic image combines canny with mathematical morphologic method is proposed.First,the filter operator constructed with the first order partial derivative of two-dimensional Gaussian function is emplyed to compute gradient with the x and y directions.Then the edge was located via the edge detection and localization method with four threshold values.And finally,edge thin operation was conducted by introducing a morphological operator.The simulation experiment demonstrates that this method will increase the precision of the image edge detection,enhance the antinoise performance,thus improving the effect of image edge detection.