Global Wheat Head Detection (GWHD) for Tanzania Using AI and Machine Learning

Local Chapter Dar Es Salaam, Tanzania Chapter

Coordinated byTanzania ,

Status: Completed

Project Duration: 31 May 2023 - 23 Jul 2023

Open Source resources available from this project

Project background.

Located in North Africa, Algeria has sought to support agriculture because of its potential in this sector. Indeed, it has put in place several agricultural policies and the objective was to achieve food security by substituting local production for imported products. One of the most important goals is the development of modern and sustainable protected greenhouse cultivation in Algeria. Managing greenhouses by means of AI technologies allows growers to be more focused on their crops and provides control at their fingertips. Proposed by the University of Ain Temouchent, Algeria, this project aims to develop ML models for the management of intelligent greenhouses

The problem.

The management of all equipment under one control system, including heating, venting, and irrigation, is a hard task in terms of systems management and data collection. As a seasonal grower with product cycles of up to two years in duration, patience is necessary. It takes time to collect the data for these systems to work and learn.

Greenhouse environments are also challenging for technology implementation due to broad temperature and humidity ranges, which influence both the electronic and mechanical components that contribute to their ongoing development. This can be a frustration for staff trying to complete their weekly plans.

AI solutions for greenhouse growers are still in their initial phases of development. The integration of intelligent control systems requires changes to processes, which can be disruptive to production, so flexibility and managing expectations are important to manage the greenhouses effectively.

Project goals.

- Build a strong community for sharing knowledge of AI and ML models in agriculture. - Decide the best values for managing the levels of temperature, humidity, the use of water, light, and other parameters.

Project plan.

  • Week 1

    Understanding the problem conducting research on best datasets and best AI approaches

  • Week 3

    Data cleaning / preprocessingn/ labelling

  • Week 4

    Model building/ choosing the best

  • Week 5

    Application building / real deployment

  • Week 6

    Visualisation and publication

Learning outcomes.

Problem solving and Hands-on real-world AI experience

Share project on: