Assessing and preventing the Impact of Desert Locust through Machine Learning

Assessing and preventing the Impact of Desert Locust through Machine Learning

  • The Results
Challenge completed! Results follow soon.

Kenya Red Cross Society is the leading humanitarian agency and the strongest humanitarian brand in Kenya. 50 technology changemakers are leveraging satellite and drone imagery for assessing the impact of desert locust in various regions.

 

The problem

Already, 3.1 million people in arid and semi-arid areas of Kenya are food insecure, and increased breeding of desert locusts, coupled with a recent flooding as well as Covid-19 pandemic, poses a wider risk of food and pasture shortage.

Regionally, Ethiopia, Kenya, Somalia, South Sudan, Sudan, and Uganda host 25.3 million people facing high levels of acute food insecurity, which is 28% of the case-load of Africa. Of these, more than 11 million people in Ethiopia, Kenya, and Somalia are located in areas currently affected by the desert locust infestations.

The following is a short sequence, recorded in Kenya in May 2020, which shows the scale of the problem.

 

 

 

The project goals

The project team will use satellite and drone imagery for a desert locust impact assessment study. The main objectives will be:

  1. Deriving accurate vegetation cover types e.g. pastureland and cropland masks from sentinel 1 and 2 imagery for the 16 counties that were adversely affected by desert locusts. This includes looking into what land cover maps exist to improve the quality of a vegetation outlook in real/near real-time.
  2. Quantifying the impact on desert locusts via drone imagery and classification models. This will help to measure the desert locust impact on croplands and pasturelands by hectares. The data comes from multispectral drone imagery captured by the WingtraOne drone.

 

This challenge is hosted with our friends at