Using CNN to identify map patches based on high-resolution data
| Autor: | Xin Zhang, Luhua Luo |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
0209 industrial biotechnology
geography geography.geographical_feature_category Land use business.industry Computer science Deep learning Feature extraction Big data Drainage basin Land-use planning 02 engineering and technology computer.software_genre Convolutional neural network 020901 industrial engineering & automation Remote sensing (archaeology) 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Land development Data mining Artificial intelligence business computer |
| Zdroj: | ICAIIC |
| Popis: | With the advent of the era of big data, remote sensing technology has continued to evolve, and the use of remote sensing technology to accurately extract land maps has also become a significant study. Nowadays, the application of machine learning in remote sensing is increasingly widespread. The use of deep learning model algorithm to extract plots intelligently has become an important means for remote sensing to obtain surface information. This study establishes an automatic extraction model based on convolutional neural network for extracting parcels from high-resolution remote sensing images. Taking the tributary of the Weigan River Basin as an example, using the high-resolution satellite data as a benchmark, the deep learning model is used to intelligently extract the cultivated land and water information in the image. Studies have shown that the use of machine learning to extract features of maps is more efficient, about five times the efficiency of artificial hand-painting. Therefore, it saves manpower, material resources, and financial resources. The study provides basic data for land use and land development, which also greatly promotes the planning process of land planning and economic construction. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|