A Hybrid Data-Driven Method for Wire Rope Surface Defect Detection

Visual inspection method (VIM) has attracted more and more attention because it is fast, nondestructive, automatic, and objective, which can replace manual inspection method or assist other non-destructive testing methods for wire ropes (WR) to a certain extent. However, it is still a challenging task to accurately detect the potential defects and identify the types from the WR surface morphology. In this paper, an efficient hybrid data-driven method based on texture features and optimized support vector machine (SVM) is proposed to solve this problem, which is called WR-IFOA-SVM. Uniform local binary pattern and gray-level co-occurrence matrix features were extracted and fused from image dataset which contains three most common states, i.e. healthy, broken and worn WRs. The inertial dynamic weight function was introduced into the fruit fly optimization algorithm (FOA) to overcome the problem that the traditional FOA cannot balance the global and local search ability. And the data mining experiments of the established feature dataset were carried out relying on the proposed WR-IFOA-SVM model, which was then compared with other methods. The experimental results show that this method can effectively detect various defect types on the WR surface, furthermore demonstrate that our method outperforms the state-of-the-art works in WR visual inspection field.