摘要

The growth and development of winter wheat at seedling directly influence the nutritional quality and the crops yield. The surface biomass reflects the crop vigor at the seeding stage. In order to monitor the winter wheat growth efficiently, accurately and at low cost, this study researched the image features of the winter wheat canopy and the best monitor model. The Daheng IMAVISION's DH-HV3151UC-M industrial camera with 3 million pixels was used to capture the winter wheat canopy images of the 20 different cultivars during the seeding stage. Then the ground fresh weight of the winter wheat was detected by using weighing method. The characteristic parameters of the image were extracted to build the wheat coverage by image processing, and the relationship between the wheat coverage and the winter wheat biomass was studied. The results showed that the winter wheat biomass was different in the different period as the vegetation, and the wheat coverage changed in accord with the winter wheat biomass of different cultivars at the different planting densities. Regression Analysis was used to establish the correlative model of wheat coverage vs. the winter wheat biomass by using 80 sets of data, and it can be found that the exponential function was the best model (R2=0.834,RMSE=0.012). The rest 39 sets of data were taken to validate the model (R2=0.777, RMSE=0.016). The result of this study indicated that it was significant for monitoring the winter wheat biomass using the canopy image at seeding.

全文