Pneumonia image detection based on convolutional neural network

Abstract

In recent years, with the continuous increase of various types of data sets and the acceleration of computer hardware, the field of artificial intelligence has achieved unprecedented development. As a core driving technology of artificial intelligence, deep learning has a significant impact on human life. With the emergence of the novel coronavirus and its rapid development into a global new infectious disease, the intelligent medical treatment of pneumonia image detection has become very important. However, due to the particularity and limitation of medical images, medical image segmentation is more challenging than other computer vision tasks. Therefore, it is urgent to carry out research on medical image segmentation. This paper takes convolutional neural network as the research direction, reasonably designs and improves the model, and improves the detection and segmentation performance of pneumonia images. The main research contents are as follows: Aiming at the problems of cumbersome steps and low accuracy in traditional image detection methods and poor detection effect and efficiency in existing deep learning methods, a detection network model based on Mask R-CNN is proposed. The experimental results show that: first, the random weighting algorithm can improve the detection accuracy and training speed of pneumonia; Second, by using ResNet with different layers, it is proved that reducing the number of layers can effectively improve the detection accuracy and training speed of pneumonia.