Research on Intelligent Identification of Pavement Diseases Based on YOLOX ShuffleNetV2 Model
In order to further quickly and conveniently process the disease images obtained by the multifunctional road condition rapid photo detection system,the YOLOX-ShuffleNetV2 neural network model is introduced into the intelligent identification of pavement diseases based on image analysis.Firstly,8 000 of the 12 500 diseased images were selected as the training set,the remaining 2 500 images were the verification set,and the remaining 2 000 images were the test set,and all images were trained,verified and tested for 2 rounds,and the results of the 2 rounds of training and testing were evaluated by using the average accuracy,the average accuracy of the whole class,the precision,the recall rate,the F1 value,and the average missed detection rate.The results show that this model has a significant effect in identifying ruts and repairing cracks,but has a poor effect on identifying cracks and pit diseases.It can be seen that the YOLOX-shuffleNetV2 neural network model can be used for intelligent identification of pavement diseases,but it is necessary to increase the sample size and improve the average accuracy.
万方数据
维普
