Brazilian Wheat Dataset

Published: 29-09-2020| Version 1 | DOI: 10.17632/3ntkg88d4d.1
Contributors:
Lincoln Schreiber,
João Gustavo Atkinson Amorim,
Adriane Parraga

Description

The experiments were conducted at the Agriculture Experimental located south of Brazil, from May to October in 2018. The study area contains several 2.5m x 1m rectangular parcels, referred to as plots, which hold two Brazilians wheat varieties. The genotypes used were TBIO Toruk and BRS Parrudo (48 Toruk plots and 40 Parrudo plots). Variability in the growth of the crops was created for all test areas, where each one received a varying quantity of nitrogen. Different Nitrogen (N) rates were chosen to generate crop growth variability, to evaluate the response of biomass and grain yield to N availability, which we called spatial variability. The database consists of images captured by Unmanned Aerial Vehicles (UAV) and biomass manual measurements. Two different processes were used to create this dataset. In the first, we collected the biomass to make a ground-truth. This step is done manually and destructively. Shoot dry biomass was determined at three growth stages: the stage of six fully expanded leaves, referred herein as V6, three nodes, and at flowering by the collecting of plants in an area of 0.27 m² for each plot. This was done to create a temporal variability. The plants collected were oven-dried at 65ºC until constant weight and weighed. Then the value is extrapolated for kilograms/hectare (Kg/ha). That is the BIOMASS value in “RGB DATA.CSV”. The second process was to acquire the images at the height of 50m above ground using a camera coupled to a DJI Matrice 100 Quadcopter. The camera is a single channel DJI X3 Visible (RGB), with 12MB resolution and 8-bit pixel depth. For agriculture acquisition, the recommended frontal overlap should be at least 80%, and the side overlap should be at least 70%. The resulting pixel size for the RGB images is 2.14cm²/px. It was a total of 30 individual images captured by the UAV using these parameters. Post-processing of the acquired images includes georeferencing and orthomosaic generation. The result can be seen in the “Orthomosaic RGB” folder, where we have eight photos from different days. At the “Orthomosaic RGB” folder, we selected from the orthomosaic regions of interest (ROI) were manually delineated from each plot, as shown by the red dots in “example_ROI.png”. The ROIs were selected in a way that excluded pixels with soil exposed. And then, from each ROI, an average of pixels and count was performed in the 3 RGB channels. This could be seen for all images in “RGB_Data_Orthomosaic.CSV”. With different days with biomass measured, at the folder “Cut Plots RGB”, we have the output for the algorithm developed who use a manually created mask to perform the segmentation of the areas of interest, i.e., the plots. There is already the identifier (id) of each plot in this mask, so there is a difference between the mask’s colors, which can be seen in figure “cut_mask.tif”. After segmentation of 88 plots for each Biomass manual cut-off day, we obtained 264 RGB images with 158x110 pixels.

Files