Remote Sensing Image Segmentation Method And Application Research For Nuclear Emergency Decision-making

When a nuclear accident occurs,nuclear emergency decision making is a key aspect to ensure public safety and mitigate the consequences of nuclear accidents,but the current nuclear emergency decision making field is facing the problem of insufficient information mining and analysis capability.To this end,this study focuses on the remote sensing image segmentation method for nuclear emergency decision making and its application,and makes full use of the information potential of remote sensing images to provide an innovative and efficient and accurate remote sensing image segmentation method for nuclear emergency decision making,so as to improve the capture capability of nuclear emergency decision elements and provide corresponding intelligent technical support for nuclear accident consequence evaluation and decision support system.The main contributions of this study can be summarized as the following two points:(1)To address the problem of insufficient extraction capability of decision elements in nuclear emergency decision support,this paper proposes an improved remote sensing image segmentation method based on U-Net network.The method improves the focus on feature regions and element information related to nuclear emergency decision making by introducing a deep-level feature extractor,an adaptive attention mechanism and a void space pyramid pooling mechanism,which in turn improves the accuracy and relevance of decision element information extraction and classification.Further,a dataset conforming to the features of nuclear emergency decision-making features is developed,and comparative analysis experiments and ablation experiments are conducted on this dataset.The experimental results show that the improved method proposed in this paper outperforms the mainstream remote sensing image segmentation network represented by Deep Lab V3 in terms of segmentation performance and computational efficiency indexes,and effectively solves the difficulty of extracting key elements such as buildings and roads in nuclear emergency decision-making scenarios.(2)In order to further verify the practical application effect and application scope of the proposed improved method in nuclear emergency decision-making,this paper develops and builds a nuclear accident wisdom decision support module based on remote sensing image segmentation method.The module can be directly applied to the nuclear accident consequence evaluation and decision support system to provide intelligent and optimized concealment routes,evacuation routes and rescue routes for personnel actions through remote sensing image segmentation and optimization calculations to achieve more accurate,scientific,reliable and intelligent decision support.It actively explores new ideas to achieve optimal decision-making and promotes technological innovation and practical development in the field of nuclear emergency decision-making.This paper realizes the segmentation and extraction of decision elements in remote sensing images by improving U-Net network for nuclear emergency decision-making,and designs and implements a nuclear accident intelligent decision support module based on the extracted decision elements and combined with intelligent optimization technology,which provides intelligent technical support for nuclear accident consequence evaluation and decision support system.