Research on Image Recognition of MURA Dataset Based on Convolutional Neural Networks
With the development of deep learning technology,automated image recognition has become one of the important requirements in many fields.Firstly,the structure and basic principles of convolutional neural network models and VGG16 network models were studied,and the image characteristics of the MURA dataset were thoroughly analyzed and preprocessed.Then,in order to improve the accuracy of image recognition,a model for image anomaly recognition process on the MURA dataset was constructed,and a parameter fine-tuning method for the VGG16 model was proposed.Through the analysis of experimental results,it can be seen that the adjusted VGG16 model has a recognition accuracy of 0.87 and a Kappa coefficient of 0.58 on the MURA dataset,which verifies the effectiveness and practicality of the improved VGG16 model.
万方数据
