Abstract: Abstract: Accurate and rapid acquisition of growth information has been one of the most important steps for winter wheat production in agricultural development and management decision-making. Most previous achievements focus on the monitoring of crop growth using Unmanned Aerial Vehicle (UAV) in recent years. Among them, the most extensive reports can be chlorophyll, biomass, plant height, and water content. Taking the winter wheat as the research object, this study aims to monitor the growth characteristics of the plant under a UAV using the Coefficient of Variation (CV) and optimized neural network. A Comprehensive Growth Monitoring Indicators (CGMICV) was also considered to integrate with the CV and different indexes, including the biomass, plant height, plant water, and chlorophyll content. In addition, the multispectral data of UAV was obtained, such as red, green, red edge, and near-infrared band. Subsequently, 16 multispectral vegetation indices were selected to analyze the correlation between the vegetation index and CGMICV, according to the characteristic band range of crops. The variance expansion factor was then calculated to screen the input variables of the model. Finally, six optimal vegetation indices were selected as the input variables of the model. As such, the growth model of winter wheat was established using the Partial Least Squares Regression (PLSR), Random Forest (RF), and Back Propagation Neural Networks (BPNN). Correspondingly, an optimal growth inversion model of winter wheat was achieved, including the determining coefficient (R2), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE), according to the combined evaluation indexes. More importantly, Genetic Algorithm (GA) was utilized to optimize the growth inversion model for the whole region. The spatial distribution of winter wheat growth was then obtained in the study area. The results showed that the correlation between the CGMICV of winter wheat was much higher than that of the single index, where the most outstanding one was the Soil Plant Analysis Development (SPAD). The best inversion model of CGMICV-BPNN was achieved for the growth of winter wheat, where a determination coefficient R2 was 0.71, and the accuracy of the model was 26.79% higher than that of the traditional one (CGMImean-BPNN), fully meeting the current accuracy for the comprehensive monitoring of crop growth. The stability of the optimized CGMICV-GA-BPNN model was significantly better than that of CGMICV -BPNN. the mean relative error median was reduced by 22.22%, and the determination coefficient R2 was also increased. The CGMICV-GA-BPNN model was then applied for the growth distribution map of winter wheat in the whole study area. More than half of winter wheat was concentrated in grade III, followed by grade I. It inferred that the overall growth of winter wheat was relatively stable. At the same time, it was also found that the optimized CGMICV -BPNN model can be used to integrate the multiple growth factors of winter wheat, indicating a better performance to quantify the growth monitoring of regional winter wheat. The findings can provide an important reference for the growth monitoring of winter wheat in crop production.