“…In terms of the methods of characterizing agricultural land use intensity, research showed that: (i) a new system has been constructed using farm accounting network data that include land use, socio-economic factors, local climate, and government subsidies to calculate the unit land cost input [ 4 ]; (ii) Polynomial regression models have been applied to detect the spatial distribution of agricultural intensity in France in order to provide an important reference for the implementation of related agricultural policies [ 5 ]; and (iii) Some scholars still use indicators of the inputs or outputs to measure agricultural intensification based on traditional agricultural statistics. Indicators of inputs include fertilizers, intercropping levels [ 6 , 7 ], nitrogen input (for arable land and permanent grassland), livestock unit density and pesticide amounts (herbicide, insecticide, and flame-retardant herbicide and insecticide) [ 8 , 9 ], while indicators of the outputs include cereal and animal husbandry products [ 10 ], which have only been applied to analyze the characteristics of agricultural land use and their informative significance, rather than focusing on the innovation of characterization methods [ 11 , 12 , 13 ]. Despite the development of remote sensing science, few scholars have used remote sensing images to characterize agricultural land use intensity dynamically and in real time [ 14 , 15 , 16 , 17 ].…”