Algorithm Details

Supported Sensors
BGNIR, Micasense, Sequoia
Supported Resolution
5 cm/pixel or less
Other Requirements

- Weeds must be visually distinguishable from crops
- Developed for use in agricultural fields

Estimated Processing Time
2 hours or less
Outputs    
Shapefile, KML



Description

The purpose of the Weed Pressure algorithm is to quantify the level of weed cover detected in each grid cell of the final output. To do this the algorithm studies 2 m x 2 m grid cells of the area of interest and quantifies the weed cover in the region. A Pressure Index ranging from 1-20 is then established for each grid cell. Percentage and area of weed cover are also calculated for each grid cell. 


This algorithm is designed to work for agricultural fields with any crop type. It should also be applicable at any growth stage of the crop, the only requirement is that weeds are not beneath the crop canopy and are still visually distinguishable from the crops.


Weeds are known to be an invasive plant that compete with other plants for space, nutrients, and sunlight. A study of Wisconsin Corn and Soybean fields found that even with common early emergence weed management techniques there was an average of about 10% yield loss in surveyed fields (Hartzler, 2009). Recognizing these weeds in the pre and/or post emergence stages of your crop can help you make more informed decisions on treating your field. 


Source: Hartzler, B. (2009). The cost of convenience: The impact of weeds on crop yields. Iowa State University. Retrieved January 19, 2018, from http://lib.dr.iastate.edu/cgi/viewcontent.cgi?article=1010&context=icm


Outputs

The Weed Pressure algorithm delivers two different output formats- Shapefile and KML. They both hold the same information in different formats. For each grid cell in the output you will have a Pressure Index ranging from 1-20, Weed area square meters, and the percentage of weeds detected. 


  • Shapefile- The shapefile output will be a 2 x 2 meter grid of the selected area. Within the attribute table of the shapefile you will find the Pressure Index, Weed area, and Weed percent for each of the grid cells. When opened there is some additional work that will need to be done to colorize the grid. This will be explained below in the Visualizing the Data section. 


  • KML- This output will have the same information as the shapefile output but in a different format. The KML format can be easily opened with Google Earth. When opened in Google Earth the 




Visualizing Your Data

The most functional format of the Weed Pressure output is the Shapefile. With some knowledge of a Remote Sensing or GIS software you are able to open and customize this output for your own use. Here we will discuss how to open the shapefile in a free GIS software called QGIS. We will also explore ways to colorize this shapefile output in the QGIS software. To download this software for free on your computer visit this link- https://download.qgis.org/ 


1. To open your Shapefile in QGIS unzip your dataset use the "Add Vector Layer" button from the QGIS Layers Toolbar. Choose the SHP format of your output and it should appear in your QGIS data viewing panel. 






2. Every shapefile layer has a "Attribute Table" that holds data on each geospatial feature in your shapefile within a table. With the Weed Pressure algorithm the attribute table will display three primary statistics on each grid cell of your area- Pressure Index, Weed Area (in square meters), Weed Percentage. 


To open the shapefile Attribute Table right click on the layer within the "Layers Panel" and select "Attribute Table"




3. You probably noticed that when the shapefile comes into QGIS all grid cells are the same color. Here we will discuss how to colorize the grid cells based on the Attribute Table values. If done correctly you are able to change the coloration of your Grid so that a range of colors represents specific values from this Attribute Table.  


To do this right click on the Layer again from the Layers Panel and select "Properties"




4.
Under the "Style" section of this screen there are options to change the coloration of your layer in many ways. First at the top of this style section select "Categorized".



5. Once this is done there will be a field called "Column" at the top of the menu. This should be set to the corresponding Column in the attribute table that you will colorize the map around. In this example we will colorize the map based on the Pressure Index readings across the area. 


6. Lower on the list there is a "Color Ramp" field. This defines the color gradient that will represent these values once the changes are applied. In the example we have chosen a Red to Green scale to represent the degree of Weed Pressure. But since we want Green to represent low Weed Pressure values (Less Weeds= Good) and Red to represent high Weed Pressure values (More Weeds= Bad) we must "Invert" the color ramp using the check box located to the right of the Color Ramp field.



7. Use the "Classify" button to categorize the values of your attribute table across the color gradient. But before applying you want to uncheck the classification for the -999 value. This represents grid cells that have No Data.




8. When you are happy with all the classification settings of the layer click "Apply" on the Properties window to see the new colorization of the data on screen.