【摘要】 针对Canny算法对噪声比较敏感,检测到的边缘不够光滑,且需要人为指定高、低阈值等问题,提出了一种自适应Canny边缘检测算法。首先,利用Canny算子的基本原理,提出了一种计算梯度的新方法;然后,根据图像梯度信息自适应地生成高、低阈值;最后,根据惯性原理进行边缘跟踪,使边缘更光滑。实验结果表明:该算法不仅能够根据图像梯度信息自适应地生成高、低阈值,还保留了原Canny算法的定位准确的优点,对噪声还具有一定的抑制作用。更多还原

【Abstract】 For the sensitivity to noise,the non-smooth detected edge and given high and lower threshold of the Canny algorithm,an adaptive Canny edge detection algorithm is proposed.First,based on the basic principle of the Canny algorithm,a new method of calculating gradient is proposed.Second,the high and lower threshold can be attained according to the image gradient information adaptively.Last,the proposed algorithm traces the edge of image according the principle of inertia,which makes the more smooth edge.The proposed algorithm can not only get adaptively high and lower threshold according to the image gradient information,but also keep the original advantages such as good location and restraint of noise.更多还原