Aiming at accurately identifying insulator defects and preventing transmission lines from insulator-damage-in-duced operational anomaly,the present work proposed a threshold-difference-based insulator defect identification algo-rithm.First the image was pre-processed and treated by adaptive threshold segmentation to achieve image binarization.In the treatment of the images,trivial areas were eliminated by morphological processing,and insulator edges were extracted using an edge detection operator.Then the defect map was extracted by substracting image information of faulty insulator from normal insulator,and the damage of insulator was determined by judging whether the size of defect pixels exceeded the threshold.After binarizing the defect map,a rectangular box was used for calibration.The experimental results indica-ted that the algorithm proposed in this paper has certain practical significance in identifying insulator defects.
万方数据
维普
