Students Applying Technology & Coding for Human Rights

CGI U 2016

(From left to right) Joe Lanzilla and Alex Luta of SATCOHR, and Febin Bellamy (GU's CGI U Campus Coordinator), represent Georgetown at the conference.

(From left to right) Joe Lanzilla and Alex Luta of SATCOHR, and Febin Bellamy (GU’s CGI U Campus Coordinator), represent Georgetown at Clinton Global Initiative University 2016 in Berkeley, CA.

Clinton Global Initiative University invited SATCOHR to its annual conference in Berkeley, California from April 1st to 3rd. Joe and I were honored to attend, represent Georgetown, and present our work. CGI U brings together students from around the world to develop commitments to action that address global challenges in the areas of education, environment and climate change, poverty alleviation, peace and human rights, and public health. Students apply online, proposing a detailed plan for their commitments to action, and the Clinton Global Initiative invites a select number of students to its annual university conference. We were excited for this excellent opportunity to present our ideas, learn about the commitments of other students, listen to fascinating speakers, attend educational workshops and office hours, and network with students and organization representatives from around the world. We learned a great deal during the conference and acquired new perspectives on our commitment and the innovative commitments of other students that addressed a wide variety of global challenges. It was a remarkable and memorable experience, and we recommend that more Hoyas apply to future conferences! We are extremely grateful to the Provost’s Office and the Mortara Center at Georgetown University for providing the funding needed for us to attend CGI U.


The SATCOHR poster that Alex and Joe presented at the conference.

The SATCOHR poster that Alex and Joe presented at the conference.

Ambiguity in Satellite Imagery

To improve the chances of satellite image admissibility in international humanitarian courts, human rights geospatial analysts, satellite technology professionals, and legal experts should work together to develop protocols for analyzing and interpreting satellite imagery. One issue such efforts can address is what types of phenomena related to human rights monitoring are most easily detected with satellite imagery and what features can be confused with them. Ambiguity in satellite images can be an obstacle to the admissibility of and weight given to this potentially life-saving form of evidence. To illustrate this ambiguity, I will use examples from “Using Geospatial Technologies to Support Human Rights Research and Documentation,” a AAAS workshop.1

While monitoring the conflict in Misurata, Libya in 2011, AAAS analysts believed that one satellite image revealed a buildup of tanks on the edge of a street in an area that experienced heavy fighting. The features they saw had the dimensions and color of camouflaged tanks, but the analysts wanted to verify this by other means. Near infrared bands, which show vegetation as bright red, along with other sources of information indicated that what appeared to be tanks were actually trees cut in a rectangular manner. One object in the street that looked like the trees in the original image, but did not appear bright red upon infrared examination, was a tank. This case study demonstrates the importance of using multiple methods of examination and corroborating satellite imagery evidence with other forms of evidence.

The rectangular objects on the side of the road have the dimensions and color of camouflaged tanks

The rectangular objects on the side of the road have the dimensions and color of camouflaged tanks.

The infrared analysis shows that what appear to be tanks are actually trees. The yellow circle indicates a tank.

The infrared analysis shows that what appear to be tanks are actually trees. The yellow circle indicates a tank.

Furthermore, in Syria the AAAS analysts found evidence of what appeared to be shell craters. However, while consulting historical imagery of the area and other information, they discovered that these features actually represented uprooted trees from an orchard. Another image showed the tracks of tractors used to remove the trees.

This image shows what appear to be shell craters.

This image shows what appear to be shell craters.

What appear to be shell craters actually represent uprooted trees. This image shows tracks of tractors used to remove the trees.

What appear to be shell craters actually represent uprooted trees. This image shows tracks of tractors used to remove the trees.

These examples illustrate potential ambiguity in satellite images. Methods such as infrared analysis and historical image comparison can help provide more information and context regarding the features present in satellite images. Documenting such methods and potential sources of ambiguity would be very helpful in creating guidelines for geospatial imagery analysis. Such guidelines may help aid the admissibility of and weight given to satellite images. For example, guidelines concerning geospatial evidence of tanks may include showing the original image with a scale used to indicate the dimensions of a tank and an infrared analysis that does not show a bright red color on the objects representing tanks. Such expert guidelines may improve the success of geospatial evidence in court. Independent geospatial human rights analysts, satellite technology professionals, and legal experts should come together to develop protocols for analyzing and interpreting satellite images. If these protocols are approved by the United Nations, this potentially life-saving technology can be effectively introduced and applied in court.


1. Wolfinbarger, Susan, Jonathan Drake, and Eric Ashcroft. “Workshop: Using Geospatial Technologies to Support Human Rights Research and Documentation.” Washington: AAAS, 09/05/2014.

MIT Researchers Develop Satellite Imagery Algorithms for International Development

A team of researchers at MIT has recently studied how satellite imagery can aid international development projects. Their work has been published in Big Data. The team developed algorithms with machine-learning techniques in order to automate the identification of houses in rural areas. They also created a system for mapping villages and determining the optimum location for the placement of solar panels. This work has been previously conducted by field teams, but the researchers’ technological advancements have improved the efficiency of the process and lowered its cost. The researchers applied their algorithms to two projects.

The first project involved choosing villages in sub-Saharan Africa for a cash grant program aimed at improving living standards in low-income rural areas. The grant agency prioritized the poorest villages, and the metric it employed to determine the poorest villages is the proportion of houses with thatched roofs to metal roofs. The agency gave higher priority to villages with a greater proportion of thatched roofs to metal ones, as the metal roofs are more expensive. The researchers expanded on their house identification algorithm by having it distinguish metal roofs from thatched ones. They were able to accomplish this by incorporating reflectance measurements, which are greater for the metal roofs.

The second project involved choosing villages in rural India for installing microgrids to provide electricity from battery-storage systems and solar panels. The researchers looked for optimum sites for the panels by finding the most efficient network configuration for distributing power. The team’s computers ran thousands of different possibilities of where to place the solar panels and distribution wires. By looking at the variations, they could determine which configurations would give electricity to the greatest number of houses with the least amount of wiring needed.

In order to develop their algorithms, the researchers had to start by manually selecting houses from satellite images to enter training data for a machine-learning system. Such a system generalizes the criteria for determining what is a house and what is not a house, so when it is shown a new image, it can recognize the houses in that image. As the system receives more examples of houses, it detects houses better. The team notes that their algorithms could potentially be used for a variety of other purposes. For example, as there is not much data on population changes in India, the algorithms could help determine which areas in India have gained or lost population over time.

Manual Identification of Houses

Manual Identification of Houses

Automated Identification of Houses with Algorithm

Automated Identification of Houses with Algorithm

The team's algorithm helped identify the most efficient locations for placing solar panels in villages (green represents the most efficient locations)

The team’s algorithm helped identify the most efficient locations for placing solar panels in villages (green represents the most efficient locations)


Shrinking Ice Caps

Geospatial work done by NASA shows that the thickest parts of the Arctic ice caps are melting at a faster rate than the thinner, younger parts. NASA scientist Josefino Comiso compiled data from NASA and DoD satellites to demonstrate this trend over several decades. Comiso found that the area of multi-year ice, or ice that has been around for at minimum two summers, is decreasing by about 17% per decade. While seasonal ice comes and goes with each summer and is meant to disappear, the gradual disappearance of multi-year ice is worrisome. Comiso constructed a time-series of multi-year ice for a period of 32 years. The satellites he got the data from used microwaves to determine the area of ice, which is a unique method of applying geospatial capabilities. The satellite used to collect the data was the Defense Meteorological Satellite Program (DMSP), which possesses the F17 Special Sensor Microwave Imager/Sounder (SSMIS). The data collected from this sensor led to the creation of the Sea Ice Index, which was funded by the National Oceanic and Atmospheric Administration (NOAA). The microwave sensors allow for observation of varying brightness levels of the ice. The brightness data are analyzed with an algorithm that distinguishes multi-year ice from other types of ice. Such an environmental issue can certainly lead to human rights issues down the line. With shrinking ice caps and rising sea levels, entire nations, like the Maldives, may end up under water.


Geospatial BBC Articles

The BBC has published a few articles this summer about advancements in satellite imaging and its applications that hold promise for the future. The first article, found here, concerns a company called PlanetLabs. PlanetLabs is a Silicon Valley start-up that makes shoe-box-sized satellites called “cubesats.” It has 28 satellites currently in orbit, and plans to release 131 in the next year. The satellites are relatively inexpensive, and with so many in orbit, human rights activists will be able to monitor areas much for efficiently and effectively. Critics say that the proliferation of satellites may open up a Pandora’s Box. The imagery may be used for corporate espionage or gaining more control over an insurgency movement. Nevertheless, PlanetLabs intends to give geospatial imagery to NGOs for humanitarian purposes.

The second article, found here, is about a US Government decision to lift restriction on higher-resolution satellite images. Until now, corporations could not take satellite images in which features smaller than 50 cm were visible. DigitalGlobe, a geospatial technology company based in Longmont, Colorado, applied to the US Department of Commerce to lift the restrictions. DigitalGlobe plans to put satellites in orbit with the capability of taking images in which features as small as 31 cm are visible. The company hopes that the images can be used to help agriculture and disaster relief. Some critics have cited concerns for privacy and national security as a result of the decision.

The last article I would like to discuss, found here, concerns a joint effort between DigitalGlobe and The Nature Conservancy to track threatening invasive plant species in Hawaii. Plants like the Australian tree fern use up water supplies in native Hawaiian forests, threatening the native flora. The images of the forests, provided by Resources Mapping, were put on a platform in which web users can search for and identify the invasive species. The platform, which has some built-in quality control, has already seen activity from thousands of users marking the locations of invasive species.

Now that more satellites are put into orbit, higher resolution geospatial imagery is allowed, and more people are interested in geospatial monitoring, the future of human rights monitoring seems very promising. One issue that is bound to attract significant discussion because of these advancements is that of privacy. How far will regulation and oversight of geospatial imagery go? What will the restrictions on geospatial imagery be? Where is the line drawn for violation of privacy? We will certainly here some of these questions discussed in the near future.


Articles Cited:

Mini-satellites send high-definition views of Earth

US lifts restrictions on more detailed satellite images

Web users join hunt for Hawaii tree invaders

Interactive Map Launched

I have added a new section to the website called “Interactive Map.” I constructed the map from the AAAS case studies found here. These case studies were conducted by the Geospatial Technologies and Human Rights Project of the AAAS. The project studies applications of geospatial technologies to human rights and determines how such technologies can be used by courts, commissions, and human rights organizations. The map is interactive, so I welcome users to document known human rights abuses with supporting evidence on the map. Please go to the map to learn more about it. The map was created with Google Maps Engine.

Using ArcGIS to Find the Victims of the Herrin Massacre

In this post I would like to write about a fascinating article I read in ArcUser Magazine. The article can be found here:

This article describes a four year effort by an interdisciplinary team to find the potter’s field of the Herrin Massacre. The 1922 massacre, which occurred in Herrin, Illinois, was the result of a labor dispute involving unionized mine workers, non-unionized mine workers, and armed guards. According to the article, in 1922 the United Mine Workers of America (UMWA) went on strike, during which W.J. Lester, owner of the Southern Illinois Coal Company, hired non-union miners and armed guards in order to resume mining operations. This led to an attack by UMWA members on the armed guards and non-union miners, with a total of 23 men killed in the massacre. The locations of the bodies of those killed was forgotten, but when Steven Di Naso, a Geographic Information Systems (GIS) scientist, and Scott Doody, a historian, decided to create an interdisciplinary team to find the bodies using GIS technology, it was not long before they made some incredible discoveries.

The team began this endeavor by using an old, hand-drawn paper map of the cemetery to build a GIS model. In order to add to this model, the team needed to compile accurate historical data. They collected and reviewed interment records, news articles, cemetery records, county recorder’s office records, and period photographs. Such resources offered geographic clues and supported hypotheses of location. The team knew that conceptual designs do not always correspond to reality, and this was the case they encountered. To fix this, they used a high-definition, high-accuracy, long-range 3D scanner to scan the topography, headstone outlines, and imagery extrapolated from millions of cloud points. They were able to process this information with ArcGIS 3D Analyst and visualize it with ArcScene. This process also suggested locations of unmarked burial sites by showing slight changes in slope and drawing attention to small surface depressions. The team also collected precise GPS coordinates for the headstones and took photographs to go along with them.

With a much more dynamic and accurate GIS model, the team was ready to use it to look for the likely burial sites of the Herrin Massacre victims. The cemetery was organized into blocks based on records, and a heat map was overlaid on the blocks to illustrate interment history. The earliest interments, in 1905, were in blocks near the center of the cemetery, and as they were not of interest, they were mostly colored violet and blue. As time passed, the interments spread to further out blocks, radiating from the cemetery’s center. The spaces with interments in 1922 were red, as they were the areas of interest. What the team discovered was that all but block 15 followed the radial pattern. Block 15 was used irregularly and its interments dispersed. Such a phenomenon is characteristic of a potter’s field, which is what the team was looking for. Based on their data, the team marked two likely burial areas for the victims within block 15.

The Herrin City Cemetery spatiotemporal model created by the team. Block 15 represents the potter's field.

The Herrin City Cemetery spatiotemporal model created by the team. Block 15 represents the potter’s field.

The team began excavations in November 2013, and they found 8 of the 12 unidentified victims from the massacre. They plan to continue excavations this summer. The graves of these victims were successfully located with GIScience and ArcGIS. The team was able to integrate thousands of historical records and other forms of data into an ArcGIS model in order to accomplish their objective. ArcGIS allowed them to process and visualize this data, which was key to finding the bodies. Such a process has great potential for human rights, where multiple forms of data can be analyzed and visualized in order to get a clear picture of the situation and allow for effective decision-making. New forms of data that could be integrated into such a model concerning human rights abuses could be social media posts of protesters and oppressed civilians, and satellite images of human rights abuses occurring. Di Naso and Doody’s team have shown the world what the power of teamwork, dedication, and GIS can accomplish.


Di Naso, Steven M., and Scott Doody. “Finding the Victims of the Herrin

Massacre.” ArcUser Spring 2014: 64-69. Print.

Geospatial Technology Applications to Human Rights Monitoring

After presenting my poster, I wanted to go into more depth with my research and wrote an essay about geospatial technology applications to human rights. This essay expanded on the computer algorithms presented in the poster, described new algorithms and their potential applications, and explained how statistics can be applied to human rights data. Lastly, I cited how a San Francisco start-up called Planet Labs is advancing the geospatial imaging landscape by putting more satellites into orbit for much lower prices. The high costs of satellite imagery and lack of satellite availability can greatly hinder human rights activists’ efforts, but Planet Labs’ endeavors give us hope for the future of human rights monitoring. Here is the essay:


Human rights activists at Non-Governmental Organizations such as the American Association for the Advancement of Science and Amnesty International have employed geospatial imaging to identify human rights abuses around the world. Often, this process involves geospatial analysts searching for human rights abuses in images with their own eyes. Computational processes and new geospatial imaging technology have great potential to aid these analysts in identifying human rights abuses more efficiently and effectively. These processes and technology include algorithms, statistical methods, and affordable satellites.

Change detection analysis is a computational process “used to determine the change between two or more time periods of a particular object of study” (Macleod & Congalton, 1998). When applied to geospatial imagery, change detection analysis methods measure differences in radiance values of the Earth’s surface (Theau, 2012). Currently, most of the applications of change detection analysis concern deforestation. In 2005, a group of scientists using the Moderate Resolution Imaging Spectroradiometer (MODIS) noticed that “once a clearing was the size of three MODIS pixels, change-detection computer programs consistently identified the areas as deforested,” which demonstrates that both advanced change detection algorithms and remote sensors must be developed to effectively track change using geospatial imagery (Morton et al, 2011).

Before change detection analysis can be applied to geospatial imagery, image pre-processing must be conducted. Image pre-processing improves images by making them appear as if they are derived from the same remote sensor (Xie et al, 2008). Common image pre-processing algorithms address radiometric errors, geometric errors, image fusion, and shadow-removal. Radiometric errors refer to differences in the brightness value of pixels (Richards, 2013).  Sources of this error are instrumentation, time of day, season, and the atmosphere. Sources of geometric errors are the velocity of the remote sensor and the Earth’s curvature (Richards, 2013). Image fusion algorithms enhance image clarity by overlaying images from different remote sensors (Xie et al, 2008). Lastly, shadow removal algorithms take into account image feature height, resolution, and the angle of the sun to remove shadows from images (Mena & Malpica). The aforementioned pre-processing algorithms enhance image quality and control for variability due to errors. In doing this they allow for more accurate change detection analysis.

Computational algorithms and statistics have already been applied to monitoring human rights abuses, and the results are promising. While conducting research at the University of Maryland, Andrew Marx utilized algorithms to identify burned villages in Darfur (Marx, 2013). Marx used images from the US Geological Survey and employed a NASA algorithm to reduce atmospheric noise, like aeorosols, and create an automatic cloud mask. Using the Interactive Data Language, Marx created an algorithm to analyze near-infrared reflectance to search for villages destroyed by burning (Marx, 2013). The algorithm picked up differences in the geospatial images between pre-burn materials, such as dead wood, and post-burn materials, such as charwood and ash.

Andrew Marx also employed statistical methods in his study to determine the stability of remotely sensed signals and accuracy of certain geospatial images (Marx, 2013). Marx conducted a one-way ANOVA to compare the variability among the images of 179 villages to the variability among the images of a single village. Ideally, there should be more variability among images of different villages than images of the same village, so a greater F-statistic suggests a more stable signal when compared to observations across all villages (Marx, 2013). Marx also used t-tests to determine the accuracy of geospatial images of the same location. He created a baseline of all images he had of a village by creating a “mean” from the images (Marx, 2013). He compared a new image, or “observation,” to this “mean.” The fewer standard deviations it was from the “mean,” the more accurate the image was (Marx, 2013).

Geospatial imaging has already demonstrated its use in monitoring human rights abuses. When combined with computational processes, statistics, and new algorithms, human rights monitoring can be even more effective. While geospatial images may take considerable time and require considerable financial resources to acquire, the geospatial technology landscape is changing. In February of 2014, a San Francisco start-up named Planet Labs released a fleet of 28 small (10 cm x 10 cm x 30 cm) satellites into space (Taylor, 2014). Planet Labs plans to release 131 more of them in the next year. They are less expensive than current satellites and will be able to provide images for analysis within a few hours of request (Taylor, 2014). With ever-advancing and more affordable technology, the future of human rights monitoring holds great promise.



Macleod, Robb D., and Russell G. Congalton. “A Quantitative Comparison of

Change-Detection Algorithms for Monitoring Eelgrass from Remotely Sensed

Data.” Photogrammetric Engineering & Remote Sensing 64.3 (1998): 207.

Marx, Andrew J. “Employing Moderate Resolution Sensors in Human Rights and

International Humanitarian Law Monitoring.” Diss. U of Maryland, 2013.

Marx, AJ. A New Approach to Detecting Mass Human Rights Violations Using Satellite

Imagery. Washington, DC: US Holocaust Memorial Museum, 2013. Print.

Mena, J. B., and J. A. Malpica. A Change Detection Method with Radiometric

Normalization and Shadows Removal in Multispectral Satellite Imagery.

Madrid: Alcala University, n.d. Print.

Morton, D.C., DeFries, R.S., Shimabukuro, Y.E., Anderson, L.O., Del Bon Espírito-Santo, F., Hansen, M., and Carroll, M. 2005. Rapid Assessment of Annual Deforestation in the Brazilian Amazon Using MODIS Data. Earth Interactions. 9: 1-22.

Richards, John A. Remote Sensing Digital Image Analysis. 5th ed. Heidelberg:

Springer, 2013. Print.

Taylor, Richard. “Mini-Satellites Send High-Definition Views of Earth.” BBC 15

May 2014, Technology: n. pag. BBC News. Web. 30 May 2014.


Theau, Jerome. “Change Detection.” Handbook of Geographic Information. Ed.

Wolfgang Kresse and David M. Danko. N.p.: Springer, 2012. 175-84. Print.

Xie, Yichun, Zongyao Sha, and Mei Yu. “Remote Sensing Imagery in Vegetation

Mapping: A Review.” Plant Ecology 1.1 (2008): 9-23. Print.


Research Poster

Organizations like AAAS and Amnesty International have undertaken huge efforts to monitor human rights conflict zones with the aid of geospatial imagery. Such images can provide valuable information about areas shut out to reporters and can be used as evidence in humanitarian courts. I researched how technologies like computer algorithms and geovisual analytics may be used to aid human rights organizations and proposed methods for incorporating them into human rights analysis methods. My findings are presented in the poster below.



For my first post, I would like to discuss the purpose of the blog and my plans for it. This blog was created on a suggestion from my faculty mentor, Dr. Ali Arab, to help facilitate my research in applications of technology to human rights and to share it with the world. This research is an undertaking I started in the spring of 2013, when I studied how the American Association for the Advancement of Science used satellite imagery to monitor human rights conflict situations around the world. I thought about and investigated how technology could help AAAS in this endeavor, and presented my results in a poster at several conferences and symposia. I will discuss and embed this poster in my next post. Since then, I have written an essay about applications of geospatial technology to monitoring human rights by expanding on the research conducted for my poster and incorporating new research. Some of the technologies I hope to investigate in the near future include volunteered geographic information, geospatial technology, algorithms, geovisual analytics, and change detection analysis. I also plan to research statistical pattern recognition and other statistical methods in relation to human rights. I welcome all communication about the blog, and I would be glad to hear your ideas and suggestions. Please feel free to contact me at or

© 2018 SATCOHR

Theme by Anders NorenUp ↑