基于改进Canny算子的绝缘子裂纹检测研究

为解决绝缘子图像中由于背景的复杂性 导致边缘检测 存在噪声及伪目标问题，提出了基于二维阈值分割的Canny算子边缘优化法。首先，在HSI颜色空间通过对S分量进行强化，有效弱化背景，突出绝缘子目标；其次，针对传统canny算子在提取绝缘子边缘时由于部分噪点突出导致检测裂纹的精度较低，提出对Canny算子中的阈值选取进行改进，采用二维直方图灰度-局部方差法对经预处理的图像进行阈值处理。所提改进算法能够有效优化绝缘子航拍图像处理结果，提高检测绝缘子裂纹准确率。

Crack Detection Method of Insulators Based on Improved Canny Operator

Targeting the problem of noise and pseudo-target in edge detection due to the background complexity of insulator image， the paper proposes a Canny operator-based edge optimization method based on two-dimensional threshold segmentation. Firstly， the S component can be strengthened in HSI color space to weaken the background and highlight the insulator target. Then in view of a lower accuracy of crack detection in extracting the edge features of the insulator with traditional Canny operator because of partial prominent noise， the threshold selection in the Canny operator is improved， and the threshold processing of the preprocessed image is carried out by using two-dimensional histogram gray-local variance method. The method can effectively optimize the results of aerial image processing and improve the accuracy of the crack detection for the insulators.