Discussion on feature enhancement technology of audit object of power transmission and transformation project settlement based on remote sensing technology

This paper proposes a feature enhancement technology for settlement audit objects of transmission and transformation projects based on remote sensing technology. Through the combination of remote sensing image preprocessing, feature enhancement and deep learning technology, various facilities and structural features in transmission and transformation projects are automatically identified and extracted. CNN is used to extract deep features of remote sensing images, and the image recognition accuracy is improved by multimodal data fusion technology. Experimental results show that this method can improve the recognition accuracy of audit objects in various environments, improve the accuracy by about 20%, and improve the processing speed by 30% compared with traditional methods. The technology proposed in this paper provides a new type of automated solution for settlement audit of transmission and transformation projects, which has wide application potential.