Due to the changeable sea current conditions and the uncertainty of the sediment situation, the piers in the offshore area are partially washed by seawater, resulting in the fragile parts of the piers being eroded by seawater. When erosion reaches a certain depth, it can threaten the safety of the bridge structure. In order to improve the accuracy of bridge pier crack identification, a bridge pier crack identification method based on an improved grasshopper optimization algorithm optimized support vector machines (SVM) is proposed. To improve the Grasshopper Optimisation Algorithm (GOA), first select piecewise chaotic functions for mapping. This will increase the diversity of GOA and improve its convergence speed. Secondly, the Cauchy mutation is added to avoid the GOA from falling into the local optimum, and the global optimum is optimized, and the adaptive update strategy is introduced at the local position update to continuously and automatically adjust the control parameter c during the iteration process, which improves the local search capability of the GOA. Finally, the improved Grasshopper Optimisation Algorithm (IGOA) is used to optimize the penalty factor and the kernel function parameter \(\sigma\) in the Support Vector Machine (SVM). IGOA is compared with various algorithms for fitness, and IGOA-SVM is compared with SVM and GOA-SVM experimentally, and the results show that IGOA has the best fitness, and IGOA-SVM is an efficient SVM optimization algorithm with obvious effects and the highest classification accuracy.