Researchers used Landsat-5 TM and Landsat-8 OLI TIRS imagery to categorise land use near Rohingya refugee camps. As of March of 2019, over 900,000 Rohingya people have been forcibly displaced to Ukhiya and Teknaf Upazilas in Bangladesh. The Rohingya people in Rakhine State, Myanmar, have suffered from discrimination, statelessness and targeted violence over the decades. Monitoring changes in the environment surrounding refugee camps, such as removal of trees, water sources and a change in the built environment, is a significant indicator of depletion of natural resources for host countries. Just as satellite imagery can help direct humanitarian aid, it can also detect the environmental impact of refugees. This information is then combined with Digital Elevation Models (3D computer graphics representation of elevation data to represent terrain) to find the location of groundwater. It uses medium-resolution satellite imagery (such as the United-States based Landsat program, and the Advanced Spaceborne Thermal Emission and Reflection Radiometer) and radar imagery to map different types of rock that is visible on the surface. Satellite imagery and auxiliary data from multiple sources can assist the exploration of underground water near refugee camps so that bores might be sunk. Updated information of land cover and land use near refugee camps can also help us understand the potential supply of crucial natural resources for daily survival such as water, firewood and building materials.
Produced refugee dwelling footprints from historical humanitarian operations based on OBIA have high potential to be used as training data for deep learning models. This allows computers to learn what a dwelling looks like from training data, with the computer developing its own criteria by which to recognise them. With the fast development of AI technology, deep-learning based approaches have become increasingly popular in extracting refugee dwellings. However, designing these rulesets requires expert knowledge and they are usually complex as the shelters are different across various regions or periods. For example, these might be areas larger than 40 square metres, rectangle shapes, bluish roof colour. The core of OBIA is to design rulesets to help the computer recognise dwellings automatically. Instead, using a semi-automated approach - what’s known as geographical object-based image analysis (OBIA) - has become popular in humanitarian operations in the last decade. However, it is tedious, time-consuming and labour-intensive because there are usually tens of thousands of dwellings to be delineated.
Satellite eyes software#
Up to now, manually delineating and checking using Geographic Information System software by experienced staff can produce the highest-accuracy dwelling footprints. Taking images of thefootprintsacross time can also help in monitoring the growth and dismantling of temporary refugee shelters.Įxperts have developed different approaches for the use of satellite imagery in extracting refugee dwellings and estimating displaced people. Updated very high spatial resolution satellite imagery such as Pléiades and WorldVie, are effective in estimating the number of people displaced. However, the building footprints, especially in remote areas, are often outdated or not available in open source data such as OpenStreetMap. To map out how much aid is required, up-to-date and high-quality data creates a set of visual representations of what refugee dwellings look like - the building ‘footprint’ - which can help estimate the size of the population in need. They struggle with a lot of challenges such as a lack of food, nutrition, clean water, sanitation, and difficult access to education or medical services. Satellite imagery can help humanitarian operations by providing immediate information on the location, number of people, the situation and environment surrounding them to help with logistics and planning.Īt the end of 2020, 82.4 million people were forcibly displaced due to persecution, conflict, violence, or human rights violations, with around 86 percent of them from countries with developing economies.