Open navigation

Field Inventory & Traits Analytics

Delair.ai provides analytics to manage inventories in crops, plantations and forestry plots. Some features are related to the vectorization of asset (field, microplots) in order to compute further analytics. This tutorial details how to run these analytics and read their results.

1. Required Inputs

This table shows what analytics shall be preliminarily calculated, plus field design information and other user input, that shall be provided to vectorize the agriculture assets, run inventories and calculate traits :

FIELD INVENTORY ANALYTICINPUTS
VECTORIZATIONSMicroplot Vectorization1. Base Scouting Maps2. Experimental design plan
Field Vectorization2. Rough field boundaries to localize field in orthomap
Row Vectorization2. Field boundaries3. Row spacing
STATISTICS BY PLOTS
2. Microplots boundaries
GAPS & COUNT
2. Field Boundaries3. Interplant spacing and canopy diameter
NB: Row Vectorization is required and generated when running Gap & Count
TRAITS
Plant Height2. DTM (bare soil)3. Optional Field boundaries
Emergence Characterization2. Field or microplots boundaries3. Gaps & Count
Flowering Characterization2. Field or microplots boundaries
Stay Green
FCOVER

Read about currently supported agriculture assets :

2. Vectorizations

2.1. Microplot Vectorization

This analytic digitalizes and georeferences microplot boundaries, to delineate microplots to then measure, benchmark and monitor traits of varieties, evaluate trial responses at microplot level. Microplot customer id is linked to each microplot to enable seamless data analysis. It serves as the reference file for traits and trial responses data storage.

2.1.1. Inputs

  • Scouting Maps shall be computed first
  • Additional Input required from the customer is the experimental design plan (.xlsx format).

All microplots must have same design. The experimental design plan shall provide the trial breakdown :

INPUTS OF EXPERIMENTAL PLANDESCRIPTION
block_row_idcustomer row id of the microplot, within a block or a field
block_col_idcustomer column id of the microplot, within a block or a field
block_plot_idcustomer id of the microplot
block_idcustomer block or field name (optional)
number of rows per microplots
number of rows of analysis
plant spacing
row spacing
row length
alley size


EXPERIMENTAL DESIGN PLAN : AN EXAMPLE


IMPORTANT - EXPERIMENTAL DESIGN PLAN : MARKING THE FIELD TRIAL START

You have to indicate in your file the location of the field trial start. In the process to run the analytic (steps detailed at 2.1.6), you will have at some point to create an annotation on the RGB map that will correspond to this field trial start position :


2.1.2. Output File Formats

The microplot customer id is linked to each microplot to enable seamless data analysis ; linkage of traits and outcomes. Output formats are :

  • .geojson
  • .shp

2.1.3. Deliverables

The output contains these fields :

FIELDDESCRIPTION
parent_idblock or field Id
block_idcustomer -block or field name
block_row_idcustomer -row id of the microplot, within a block or a field
block_col_idcustomer column id of the microplot, within a block or a field
block_plot_idcustomer id of the microplot

2.1.4. Display

From layers left panel, select Microplot layer from SURVEY DATA section, inside Contours subgroup :

Outputs (id's) can be browsed from the information panel on right side :

2.1.5. How to run the analytic ?

Microplot vectorization can be run using either :

  • A field trial experimental plan from an excel file (standard workflow)
  • or a shapefile containing the microplots (in this alternative, it is highly recommended to also upload an experimental plan if any, as some adjustments may be required)

2.1.5.1. Standard Workflow

Step 1/9 - Ensure to have the base scouting maps processing completed in your site.

Step 2/9 - Open your site and start to upload of your experimental plan from UPLOAD FILES button on top bar :

Step 3/9 - Select to attach the experimental plan to All surveys of this site ; this will mount the experimental plan as a reference file :

Step 4/9 - Drag-and-drop or browse your experimental plan file :

Step 5/9 - Edit file Name if necessary, select Type : File and Category : Microplots / Plots. Then press UPLOAD :

Your experimental plan will be stored and made available under Downloads section :

Step 6/9 - Display the RGB layer from the layers panel at BASE LAYERS section. Create on this RGB map a point annotation to mark the position of the field trial start. Reminder : the field trial start should also have been marked in your experimental plan excel file (cf. 2.1.2.). You can optionally modify the annotation icon from the STYLE section :

Step 7/9 - Rename this point annotation to "Start of field trial" :

Step 8/9 - Go to Analytics section from the left panel, scroll to find Microplot Vectorization section, and press Run :

Step 9/9 - Select your experimental plan among the listed reference files then press Launch microplot vectorization button :

Once process is complete vectorization results becomes viewable as shown at section 2.1.5.

2.1.5.2. Alternative Workflow using a shapefile

This section will be documented soon.

2.2. Field Vectorization

This analytic digitizes the exact contour of a cropped area within a field, to know precisely the geolocation of field boundaries and its cropped surface.

2.2.1. Inputs

  • Scouting Maps shall be computed first
  • Field boundaries must be distinguishable from the sky (tracks, roads, alleys, bare soil). Inputs to provide are the rough boundaries of the field, to know where is the field in the orthomosaic.

2.2.2. Output File Formats

  • .geojson
  • .shp
  • .csv

2.2.3. Deliverables

The field boundaries with the field area as an attribute.

2.2.4. Display

Open SURVEY DATA section from the layers panel and enable Field boundaries from the Contours section :

2.3. Row Vectorization

This analytic automatically identifies rows within fields in order to assist with field counting or help guide farming equipment. It detects, digitizes and measures rows.

2.3.1. Inputs

This analytic is available for crops planted in row and where bare soil is visible in between the rows. Base scouting maps have to be computed first, theoretical field boundaries and row spacing must be provided.

2.3.2. Outputs File Formats

  • .geojson
  • .shp
  • .csv

2.3.3. Deliverables

Rows digitization with the following attributes :

  • row id
  • parent id (microplots) and vegetation length in case of microplots

2.3.4. Display

Open Contours belonging to SURVEY DATA section on left panel. Click any row on map to open its information panel on right side and read row id and parent id :

3. Statistics by plots

This analytic automatically extracts statistics metrics of scouting maps and traits at the microplot or plot level.

3.1. Inputs

The basic scouting maps have to been computed first, plus the microplots boundaries defined. Quantitative scouting maps, as CIR and RGB are out of scope.

3.2. Output File Formats

  • .geojson
  • .shp
  • .mbtiles
  • .csv

3.3. Deliverables

Statistics per microplots or plots are delivered for each selected scouting maps with the following attributes :

  • min
  • max
  • average
  • standard deviation 
  • variance

3.4. Display

Enable the statistic to be displayed from the Layers panel, and open right panel displaying properties.

4. Gaps & Count

This analytic automatically determines plant count and gaps through detection.

4.1. Inputs

  • Base scouting maps
  • Field or microplots boundaries
  • Rows for row crops and orchard and vineyards planted in rows
  • Interplant spacing
  • Canopy diameter (from above) 

It is mandatory to see bare ground in between two plants on the map, otherwise only visible gaps will be detected.

For row crops, with 6cm of GSD and a multispectral camera, optimal corn stage will be 4 leaves, and starbud for sunflower.

With enhanced resolution, earlier stages can be targeted. The basic rule is : size of GSD <= half canopy size (from above) 

For orchards and vineyards plants must be planted in rows. If canopy is closed, only gaps will be detected. Height of the plant must be > 50cm.

For palm trees, plant must be > 50 cm as well.

4.2. Output File Formats

  • .geojson
  • .shp
  • .csv

4.3. Deliverables

Gaps are only for plants in rows and are delivered with the following attributes : 

  • Gap length
  • Position of gaps at line end or not

Plant count are delivered with the following attributes:

  • Number of plant per microplot or per field
  • Number of missing plant per microplot or per field

4.4. Display

From layers panel, select Inventory layer from SURVEY DATA section , and select which count or gaps layers to display :

Opening the layer information panel on right side enables to modifiy the view for certain attributes. In example below we discriminate gaps that are inside the rows or located at their ends :

5. Traits

5.1. Plant Height

This analytic automatically estimates plant height for phenotyping or volume estimation in forestry.

5.1.1. Inputs

  • Basic scouting maps (DSM and NDVI if no DTM)
  • Optional DTM (bare soil)
  • If no bare soil survey is already available, an area of bare soil must be visible in the survey (such as alleys or tracks).
  • Field boundaries

Plants should be taller than 50 cm ; overlap and sidelap shall be of minimum 80%

5.1.2. Output File Formats

  • .tif
  • .geojson
  • .shp
  • .mbtiles
  • .csv

5.1.3. Deliverables

  • Vegetation height map
  • Extrapolated DTM if no bare soil option selected
  • Statistics at field / microplots level if option selected

5.1.4. Display

Open Inventory then Plant height subgroups of layers belonging to SURVEY DATA section on left panel. Enable Vegetation height layer, click it to open its information panel on right side to read distribution or show a value in any point of the map :

5.2. Emergence Characterization

This analytic automatically calculates the percentage of green (or leaves) to characterize seedling vigor — suitable for early stage crops. It is available for crops planted in rows, where bare soil is visible between the rows.

5.2.1. Inputs

5.2.2. Output File Formats

  • .tif
  • .geojson
  • .shp
  • .csv

5.2.3. Deliverables

  • Emergence mask
  • Emergence % per microplot

5.2.4. Display

Open Inventory then Emergence characterization subgroups of layers belonging to SURVEY DATA section on left panel. Emergence % per microplot is displayed under Emergence ratio layer, click it to open its information panel on right side to read value chart :

The Emergence mask belongs to the same layers subgroup :

5.3. Flowering Characterization

This analytic automatically calculates the percentage of flowering vs. no-flowering — currently suitable for row crops with yellow flowers only such as canola and sunflower.

5.3.1. Inputs

  • Basic scouting maps (RGB Bird View)
  • Boundaries of the field or microplots
  • Yellow flowers shall be visible from the sky

5.3.2. Output File Formats

  • .tif
  • .geojson
  • .shp
  • .csv

5.3.3. Deliverables

  • Flowering mask
  • Flowering % per microplot

5.3.4. Display

Open Inventory then Flowering characterization subgroups of layers belonging to SURVEY DATA section on left panel. Flowering % per microplot is displayed under Flowering % layer, click it to open its information panel on right side to read value chart :

Flowering mask is under the same layer subgroup ; Value at point can notably be used to read value in any point of the map.

5.4. Stay Green

This analytic provides an automatic assessment of the ability for the crop to remain as a green growing plant late in the season.

5.4.1. Inputs

5.4.2. Output File Formats

  • .tif
  • .geojson
  • .shp
  • .csv

5.4.3. Deliverables

  • Stay green mask
  • Stay green % per microplot

5.4.4. Display

Open Inventory then Stay green subgroups of layers belonging to SURVEY DATA section on left panel. Stay green % per microplot is displayed under Stay green % layer, click it to open its information panel on right side to read value chart :

Stay green mask is under the same layer subgroup ; Value at point can notably be used to read value in any point of the map.

5.5. FCOVER

FCOVER, standing for Fraction of green vegetation COVER, is a measur of the fraction of ground covered by green vegetation. It quantifies the ratio between biomass and soil, i.e. the spatial extent of the vegetation. Values are between these two extrema :

0 = only soil

1 = we only see biomass

5.5.1. Inputs

5.5.2. Output File Formats

  • .tif
  • .geojson
  • .shp
  • .csv

5.5.3. Deliverables

  • FCOVER per microplot (Fcover %)
  • Biomass mask (Fcover mask)

5.5.4. Display

Open Inventory then Fraction of vegetation cover subgroups of layers belonging to SURVEY DATA section on left panel. FCOVER per microplot is displayed under Fcover % layer, click it to open its information panel on right side to read value chart :

Biomass mask is under the same layer subgroup ; Value at point can notably be used to read value in any point of the map.

Learn more with Weed Management Analytics.

Did you find it helpful? Yes No

Send feedback
Sorry we couldn't be helpful. Help us improve this article with your feedback.