Image Recognition of Pile Foundation Concrete Cracks in Salt-Photovoltaic Complementary Photovoltaic Power Station Based on CS-BP Algorithm

The generalization ability in the process of identifying pile foundation concrete cracks in salt-photovoltaic complementary photovoltaic power stations is poor, and there is an intersection between the global optimization ability and the nonlinear mapping ability, which leads to low accuracy in concrete crack image recognition. To address this issue, an image recognition method for pile foundation concrete cracks in salt-photovoltaic complementary photovoltaic power stations based on the CS-BP algorithm is proposed. Processes such as median filtering denoising, Teager energy operator enhancement, and data fusion are designed. Based on the CS-BP algorithm, combining the global optimization ability of the cuckoo search algorithm and the nonlinear mapping ability of the BP neural network, the pile foundation cracks in the salt-photovoltaic complementary photovoltaic power station are identified. The parameters are optimized through the backpropagation algorithm, and the relative threshold technology is introduced to define the crack boundary, obtaining the scale information of the pile foundation concrete cracks in the salt-photovoltaic complementary photovoltaic power station. The experimental results show that the research method can significantly improve the recognition accuracy of crack images. Meanwhile, it shows a lower proportion of false positives and false negatives in the confusion matrix evaluation, effectively improving the recognition accuracy of pile foundation concrete crack images in salt-photovoltaic complementary photovoltaic power stations.