Landscape Monitoring

  • This page describes the data provided for landscape monitoring, including crop monitoring history and feature properties.

  • For landscape features corresponding to fields, a crop monitoring history provides the most likely crops grown over periods of time.

  • The GeoJSON Feature Object properties for MonitoredLandscape data include ID, landscape type, area, class confidence, capture timestamp, and monitoring prediction.

  • The monitoring prediction property for fields contains predictions organized by start and end timestamps, along with the top three most likely crops and their confidence levels.

  • Supported crops for monitoring predictions include Bajra, Chilli, Corn, Cotton, Gram, Groundnut, Mustard, Rice, Sorghum, Soybeans, Sugarcane, and Wheat.

This page describes the data provided for landscape monitoring, including crop monitoring history and feature properties.

Crop Monitoring History

For landscape features corresponding to fields, we provide a crop monitoring history at the feature level. This history contains a list of entries where each entry describes the most likely crops growing during a finite period of time that the entry corresponds to (typically one crop season or shorter).

Feature property

This table describes the properties of a GeoJSON Feature Object from the MonitoredLandscape data.

Property Type Description
id string A feature ID, represented by the plus code of the centroid of the feature.
properties.alu_type string The landscape type of the feature. Monitoring is restricted to the field class.
properties.area_sq_m number The area of feature in square meters.
properties.class_confidence number The confidence in the accuracy of the landscape segmentation. The value is between 0 and 1.
properties.capture_timestamp_sec integer The capture timestamp in seconds for the segmentation observation's source image.
properties.monitoring_prediction array[object] The landscape monitoring history for the feature. Each entry in the history itself is a nested object. This is discussed in the next section.

Monitoring Prediction property

The predictions for fields, that is properties.monitoring_prediction are organized as follows:

Property Type Description
start_timestamp_sec integer Start time for the prediction, in seconds. Notional of start of the detected agricultural season.
end_timestamp_sec integer End time for the prediction, in seconds. Notional of end of the detected agricultural season.
crop_1 string Name of the most likely crop.
conf_1 number Confidence of the most likely crop, in [0,1].
crop_2 string Name of the second most likely crop.
conf_2 number Confidence of the second most likely crop, in [0,1].
crop_3 string Name of the third most likely crop.
conf_3 number Confidence of the third most likely crop, in [0,1].

Supported crops include: Bajra, Chilli, Corn, Cotton, Gram (Bengal gram and Chickpeas), Groundnut, Mustard, Rice, Sorghum, Soybeans, Sugarcane, and Wheat.