References | Application | Algo | Target | Data type | Data pre-processing | Extracted features |
---|---|---|---|---|---|---|
[31] | Crop fieldsmapping | RF | / | SatelliteDigitalGlobe Worldview-2 | Hand digitisation | Randomised Quasi-Exhaustive features |
[41] | Crop mapping | Decision tree | Soybean | Satellite | Multi resolutionsegmentation | NDVI NIR (near infrared) SWIR (short wave infrared) |
[131] | Crops mapping | RF | / | GF-1 WFV sensorsatellite images | Multi-resolutionsegmentation | temporal, spectral textural features vegetation indexes(NDVI, EVI...) |
[22] | Crop fieldsmapping | RF | Paddy rice | Satellite images | Polarisation for cloud contamination by Google Earth Engine | NDVI, EVIland surface water index LSWI |
[115] | Crop mapping | Deep learning: autoencoder CNN, Full CNN | Soybean, maize cotton | Satellite images | Data were pre-processed | Texturepixel’s features based on the image patch |
[157] | Crop classification | LSTM | / | Satellite & opticalimagesfield surveys | Segmentaionpan-sharpening and mosaic of optical imagesthermal noise removal and radiometric correction | Spatial features |
[76] | Crop classification | Deep learning CNN | Wheat, maize sunflower soybeans, sugar beet | Satellite images | segmentation and data restoration using unsupervised NN self-organising Kohonen maps) | Spectral and spatial features |
[33] | Plant classification | Deep learningCNN | 22 plants | Camera and cell phone images | Data are not pre-processed | Self-learned features |
[26] | Crop classification | Ensemble learningANN, DT, SVM | Rice, soybean, corn cotton | Remote sensing images | USGS online system, used a cubic convolution 245 re-sampling and a standard terrain correction incorporating ground truth points | NDVI, levels of greenmoisture |
[100] | Crop classification | set of classifiers SVM(RBF kernel), RF, Spectral Angle Mapper | Tree crops, sugar beet alfalfa, cereals | Sensor satellite Time series and images | Atmospheric correction and Radiometric calibration and Pan-Sharpening | NDVI |