A Real-Time Bridge Crack Detection Method Based on an Improved Inception-Resnet-v2 Structure

Bridge crack detection is essential to ensure bridge safety. The introduction of deep learning technology has made it possible detect cracks automatically and accurately. In this study, the Inception-Resnet-v2 algorithm was systematically improved applied real-time cracks. We propose an end-to-end model based on a convolutional neural network. This combines advantages Inception convolution residual networks, broadening network width alleviating training problem calculation speed while still ensuring accuracy. Multi-scale feature fusion enables extract contextual information different scales, which improves accuracy recognition. GKA (K-means clustering method genetic algorithm) realizes accurate segmentation target area, greatly enhances effect, effectively speed. model, large fracture datasets are used for testing without pre-training. experimental results show that performance in all aspects: accuracy, 99.24%; recall, 99.03%; F-measure, 98.79%; FPS(Frames Per Second), 196.