索贝尔算子

罗伯特算子、高斯拉普拉斯算子、索贝尔算子、蒲瑞维特算子被用来检测边缘。

Roberts operator , Log operator , Sobel operator , Prewitt operator are used to detect edges .

索贝尔算子简单有效，坎尼算法和高斯的拉普拉斯算法能产生较细的边缘。

The analyses indicate that the Sobel operator is more simple and effective , the Canny detector and the Gausian 's Laplacian algorithm yield thinner edges .

通过实验将此算法和普适性的图像识别算法（索贝尔算子边缘识别、哈夫变换）进行结果对比，证明此算法运算速度快、识别结果准确，同时具备较强的抗噪声能力。

After experimental comparison between auther 's recognition algorithm and normally used algorithms ( So-bel algorithm operator , Hough transformation , ect , ), the result shows that this novel algorithm can recognize the pattern accurately and quickly , as well as less affected by noise .

在利用改进的SUSAN算子进行亚像素角点检测时，综合应用了索贝尔边缘算子、灰度平方重心法等方法。

Sobel operator and gray square centrobaric arithmetic were synthesized with the improved SUSAN operator in subpixel corner detection .

此方法用于标准图像的分割中，选用5×5亚像素数目提取的最奇异性集合与索贝尔（Sobel）算子（默认阈值为36.7920）提取的边缘的峰值信噪比为9.3981dB。

Comparing most singular manifold extracted from 5 × 5 sub-pixel method and edge from Sobel operator ( default threshold : 36.7920 ), the peak signal-to-noise ratio is 9.3981 dB , when the algorithm is used in standard image segmentation .