Research On Vision-Based Defect Online Detection System Of Insulator Core Rod

With the rapid development of economic construction, Super High-voltage transmition lines are playing a more and more important role. Composite insulators are used to support the guide wire and also used for insulation in the transmition lines. Because of its hydrophobicity, high structural strength and light weight, composite insulators have gradually replaced the traditional ceramics and glass insulators. As composite insulators play such a key role in grid safety, the quality assurance of the composite insulators, especially the insulator core rod which is the most important part, becomes quite essential.This thesis mainly studied the key technologies in vision-based online defect detection system of insulator core rod. It aims at improve the efficiency and effect of the insulator core rod defect detection.The content of this research includes:image extraction, feature extraction and analysis, classification algorithm, and system implementa.In this research,Firstly, a defect extraction algorithm is proposed. Based on the study of the characteristics of insulator core rod, a defect extraction algorithm is designed. By target marking, defects location, noise reduction and defects merge, a good defect extraction is achieved.Then, the feature extraction and analysis are accomplished.18gray features,20projection features and16texture features are extracted. Through principal components analysis, feature vector with8elements is obtained.After that, the automatic defect classification method is studied. Based on the characteristics of the experiment condition, SVM is chosen as the basic algorithm. After a comparative analysis,21SVM models are constructed to compose the classifier through one-against-one method, and the parameter selection is also discussed. The experiment shows a good result with98%average accuracy.Finally, the prototype of the defect inspection software is build. The main function modules of the system is accomplished standing on the analyzing of the requirement.In conclusion, the key technologies of the visual-based online defect detection system of insulator core rod are studied and implemented in this paper, and a good experiment result is achieved.