Research on Bridge Pier Crack Feature Image Recognition and Classification Method Based on IGOA-SVM

In response to the issues of low automation, strong subjectivity in detection results, and serious impacts on the safety of bridges and public outings caused by existing methods for detecting cracks in bridge piers, this paper proposes an improved grasshopper optimization algorithm (IGOA) to optimize a support vector machine (SVM) classification model for crack detection. First, feature extraction of crack images is performed using a threshold segmentation method. Then, interference information such as surface impurities is removed using a binarization inverse operation and noise removal algorithm. The IGOA-SVM model is then employed to identify and classify the crack images of bridge piers. Finally, the accuracy and precision of the classification results are analyzed and compared among the SVM, grasshopper optimization algorithm (GOA)-SVM, and IGOA-SVM models. The results indicate that the proposed IGOA-SVM model improves both accuracy and precision, providing a new approach for the accurate identification of cracks in bridge piers.