References | V1 | V2 | V3 | V4 | ML | Complexity | Device | Task |
---|---|---|---|---|---|---|---|---|
[2] | 224 images | / | No | Image Data digital images | SVM | \(O(n^2*P+n^3 )+O(n_{sv}*P)\) | Digital camera | Classification |
[18] | 3*2 years of data monitoring | 1 year | / | Sensor data: soil properties | FCM | time: \(O(n*d*c^2*i)\) space:\(O(n*d + n*c)\) | Pressure-based AgLeader | Clustering |
[26] | / | / | No | Satellite data: Images in GeoTiff | EnsembleLearning (DT+ SVM+ ANN) | \(O(n^2*P)+O(P)+ O(n^2*P+n^3 )\) \(+O(n_{sv}*P)+\) \(O(ep*n(nl_{1}*nl_{2}+ nl_{2}*nl_{3}+...)+\) \(O(P*nl_1+nl_1*nl_2+ nl_2*nl_3+...)\) | Satellite | Classification |
[38] | 87.8K | / | Yes | Image data: Open database images | CNN | \(O(T*Q*t*q)\) | / | Classification |
[40] | / | / | Yes | All data types: yield climate informationsoil Geo-physical NDVI Remote sensed | RF | \(O(n^2*P * n_{trees} )+O(P * n_{trees} )\) | Yield monitor soil-maps, EM gamma survey MODIS | Prediction |
[46] | 6217 | / | / | Historicalsensor data: Yield climate | SVR, KNN, ANN | \(O(n^2P+n^3 )+O(n_{sv}*P);\) \(O(n*P);\) \(O(ep*n(nl_{1}*nl_{2}+ nl_{2}*nl_{3}+...)+\) \(O(P*nl_1+nl_1*nl_2+ nl_2*nl_3+...)\) | Spriter-GIS system | Prediction |
[57] | 229 | 1 year | Yes | Historical data: Crop yield | K-means | \(O(n*c*d*i)\) | / | Clustering |
[59] | Precipitation: 47554 min/max temperature:24542 mean temperature:14835 | / | Yes | Sensor data: Crop yield soil Biophysical climate water photo-period, fertilisation | RF | \(O(n^2*P*n_{trees} )+O(P*n_{trees} )\) | / | Prediction |
[33] | 10413 | / | / | Image data: Digital images | CNN | \(O(T*Q*t*q)\) | Cell phone | Classifiation |
[73] | / | 1 year | No | Historical data: Crop yield soil parameters | ELM | \(O(L^3+L^2*n)\) | / | Prediction |
[75] | 96 | 1year | Yes | Image data: Digital images | SVM, ANN, NBKNN DT Discriminant analysis | Discriminant analysis: \(O(n*P^2 )\) NB: \(O(n*p)+ O(P)\) | Camera Nikon CoolpixL22 | Classification |
[76] | 4 Landsat-8 scenes 15 Sentinel-1 scenes | / | Yes | Satellite data: Multi-temporal multi-source images | CNN | \(O(T*Q*t*q)\) | Landsat-8, Sentinel-1A satellites | Classification |
[152] | 8945 | multi-spectral image: 8 days interval for 30 per year | Yes | Satellite, sensor data: surface reflectance land surface temperature land cover | Gaussian CNN | \(O(T*Q*t*2)\) | MODIS satellite | Prediction |
[62] | / | 2006-2011 8 days period 32 times | / | Satellite sensor data: NDVI Precipitation land surface temperature | Rulequest cubist | \(O(n^2*P)\) | MODIS satellite | Prediction |