The wheat above-ground biomass (AGB) is an important index that shows the life activity of vegetation, which is of great significance for wheat growth monitoring and yield prediction. Traditional biomass estimation methods specifically include sample surveys and harvesting statistics. Although these methods have high estimation accuracy, they are time-consuming, destructive, and difficult to implement to monitor the biomass at a large scale. The main objective of this study is to optimize the traditional remote sensing methods to estimate the wheat AGB based on improved convolutional features (CFs). Low-cost unmanned aerial vehicles (UAV) were used as the main data acquisition equipment. This study acquired image data acquired by RGB camera (RGB) and multi-spectral (MS) image data of the wheat population canopy for two wheat varieties and five key growth stages. Then, field measurements were conducted to obtain the actual wheat biomass data for validation. Based on the remote sensing indices (RSIs), structural features (SFs), and CFs, this study proposed a new feature named AUR-50 (multi-source combination based on convolutional feature optimization) to estimate the wheat AGB. The results show that AUR-50 could estimate the wheat AGB more accurately than RSIs and SFs, and the average R² exceeded 0.77. In the overwintering period, AUR-50_MS (multi-source combination with convolutional feature optimization using multispectral imagery) had the highest estimation accuracy (R² of 0.88). In addition, AUR-50 reduced the effect of the vegetation index saturation on the biomass estimation accuracy by adding CFs, where the highest R² was 0.69 at the flowering stage. The results of this study provide an effective method to evaluate the AGB in wheat with high throughput and a research reference for the phenotypic parameters of other crops.

In order to realize the rapid and non-destructive monitoring of wheat biomass, field wheat trials were conducted based on different densities, nitrogen fertilizers and varieties, and unmanned aerial vehicle (UAV) was used to obtain RGB images in the pre-wintering stage, jointing stage, booting stage and flowering stage of wheat. The color and texture feature indices of wheat were obtained using image processing, and wheat biomass was obtained by manual field sampling in the same period. Then the relationship between different color and texture feature indices and wheat biomass was analyzed to select the suitable feature index for wheat biomass estimation. The results showed that there was a high correlation between image color index and wheat biomass in different stages, the values of r were between 0.463 and 0.911 (P < 0.05). However, the correlation between image texture feature index and wheat biomass was poor, only 5 index values reached significant or extremely significant correlation level. Based on the above results, the color indices with the highest correlation to wheat biomass or the combining indices of color and texture features in different growth stages were used to construct estimation model of wheat biomass. The models were validated using independently measured biomass data, and the correlation between simulated and measured values reached the extremely significant level (P < 0.01), and root mean square error (RMSE) was smaller. The R² of color index model in the four stages were 0.538, 0.631, 0.708 and 0.464, and RMSE were 27.88, 516.99, 868.26 and 1539.81 kg/ha, respectively. The R² of the model combined with color and texture index were 0.571, 0.658, 0.753 and 0.515, and RMSE were 25.49, 443.20, 816.25 and 1396.97 kg/ha, respectively. This indicated that the estimated results using the models were reliable and accurate. It also showed that the estimation models of wheat biomass combined with color and texture feature indices of UAV images were better than the single color index models.