Population data provides statistical information that can in turn support decision making processes. It provides essential information regarding many practices such as rescue operations or humanitarian actions, which require the estimation of local population. While traditional approaches are possible, they tend to be time consuming and expensive. Copernicus Earth Observation data (Sentinel 2 satellite images) provides high resolution satellite imagery that could be a useful alternative to collecting census data since it is signi cantly cheaper than traditional methods. The purpose of the following study examines a population extraction method that is based on open source high resolution images retrieved from Sentinel 2, as proposed by the remote task[1] by the imageCLEF2017[2] campaign. This paper describes a methodology for exporting a population estimation using satellite imagery based on the use of classi cation techniques coupled with a statistical forecast on historical data.
The work described in this paper is directed at the automated estimation of population using Supervised Classi cation through a GIS platform applied to satellite imagery. The motivation for the work is that the collection of population data, when conducted in the traditional manner can be time consuming and ine cient. This is especially the case in rural areas that lack of sophisticated communication and transport infrastructure, especially in developing countries. The solution proposed in this paper is based on the concept of using satellite imagery in order to estimate the population. The concept includes using a small sample of satellite images to build a classi er that can then be used to predict populated areas. The main issue to be addressed is how to best use the classi ed result towards matching it to the number of people in the study area. In order to correspond to the task described above, the following methodology is proposing (i) an image supervised classi cation technique , (ii) a statistical forecasting procedure and (iii) a regional disaggregation technique which will be furtherer explained in the following sections. 2
With this chapter, the methodology of downloading and analyzing Sentinel2
data shall be described. The Semi-Automatic Classi cation Plug-in (SCP)[
In order to gather the Satellite imagery, The Copernicus Open Access Hub can be used, which provides complete, free and open access to Sentinel-1, Sentinel2 and Sentinel-3 user products. The SCP Sentinel-2 tab allows the searching and downloading of the desired imagery (Level-1C) using the Data Hub API. The search area is de ned accordingly for each one of the areas of interest to include all of the respective subregions. For the purposes of this study, the date of acquisition was de ned as of 2016 or later in order to make the classi cation process relevant to the desired results. The maximum cloud cover shall be set not to exceed 10 % to help with the process whereas bands 1, 9, and 10 shall not be included. 2.2
In this section, the steps followed when classifying the satellite images will be explained. For the purposes of this study, several compounds of Uganda and Lusaka are examined. The two regions are classi ed separately. The same approach is followed for both of the study regions. It is noted that multiple di erent satellite images might be necessary to cover the entire study area. For the Sentinel-2 images, the bands are converted to re ectance and clipped to an extend large enough to cover each area examined so as to reduce the computational time of the process. The band set which is the input image for SCP is de ned and the regions of interest are created.
Iterating through the di erent color composites, allows completion of the classi cation. The de ned buildings macro-class is the area of interest that will be extracted. This shall provide us with high resolution two colored pixelated images that hold information regarding the buildings of the areas. Such data shall be used for completing the methodology through the process that will be furtherer explained in the next sections. The classi cation results can be validated through comparison with the original raster les as well as old shape- le layers provided by various open data sources (e.g. OpenLayers Plug-in, Geofabrik).
Following the completion of the process described above, the two classi ed
regions are separated using polygons that describe the sub-areas of interest. For
the purposes of this study, the PNG images of the clipped images (17 for Uganda,
89 for Lusaka) are extracted, each one now containing a certain number of blank
pixels corresponding to the living areas of each district. Concerning their usage
in the estimation algorithm that will be described, it is important to note that
all extracted classi ed les of each study area shall correctly correspond to the
respective area's land coverage ratio.
While the results of the classi cation process can provide a decent overview of
the population distribution, the buildings information shall be used in a way
that can allow to extract a population estimation. The historical Census data of
the countries can be used to predict the aggregated population result that can be
later split into the study areas accordingly through the classi ed imagery. The
statistical analysis of the Census data consists of two basic steps, (i) constructing
the population growth time-series and (ii) a forecasting model in order to pertain
to the out of sample observation; the current year population number.
Various Census data for demographic purposes are used for both areas [
A cross-validation competition between the methods Naive, MAPA[
ETS -4,143.2 -0.01 % 0.31 % 20,334,332,759 0.31 % ARIMA -29,783.84 -0.08 % 0.09 % 14,403,091,721 0.09 % SES 701,607.45 2.01 % 3.96 % 1,816,289,360,487 3.93 %
Holt -4,140.01 -0.01 % 0.31 % 20,334,001,686 0.31 % Damped 1,361.34 0.01 % 0.3 % 20,354,628,986 0.3 % MAPA -4,683,085.6 -17.93 % 19.79 % 42,915,827,697,557 17.12 % Theta 132245.72 0.91 % 1.5 % 50,440,089,294 1.5 % Naive 335,253.88 2.53 % 2.53 % 125,691,193,414 2.56%
ETS -9,335.3 -0.06 % 0.3 % 6,423,231,745 0.3 % ARIMA -6,156.78 -0.02 % 0.21 % 5,608,027,543 0.21 % SES 256,146.93 1.77 % 3.28 % 232,161,298,596 3.27 %
Holt -9,334.91 -0.06 % 0.3 % 6,423,208,336 0.3 % Damped -5,292.38 -0.03 % 0.3 % 6,161,968,390 0.3 % MAPA -1,742,891.55 -15.24 % 16.82 % 5,825,850,536,317 14.89 % The results extracted are 40,989,197 and 16,932,235 for the two regions respectively. 2001 24,146,152 10,723,229 2002 24,945,231 11,000,608 2003 25,786,777 11,282,992 2004 26,666,251 11,575,821 2005 27,578,578 11,884,613 2006 28,521,669 12,212,550 2007 29,496,442 12,560,093 2008 30,503,193 12,926,628 2009 31,540,776 13,311,214 2010 32,608,271 13,712,644 2011 33,704,880 14,130,483 2012 34,830,481 14,565,054 2013 35,987,004 15,016,334 2014 37,178,179 15,483,715 2015 38,407,677 15,966,555 2016 39,675,318 16,418,477 2017 40,983,129 16,885,858 3.3
The processes described above provide with an arithmetic output regarding Uganda's and Zambia's total population. Such information is used to proceed to the following step which is a top-down division of the countries' population in order to come to a result about each speci c area of interest in the given shapeles. It is however important to note that the subareas of interest might not add up to the entire countries' extend and therefore a straightforward division of the entire population to the acreage of the living areas (as obtained from the classi cation process) might not be possible. Because of this di culty, a di erent approach shall be examined. In order to split the total population into the di erent areas, a weight variable will be calculated for each pixel corresponding to living areas in the classi ed images. To achieve this, we gather Census data about housing and population in a small subregion of each one of the countries of interest and estimated the multiplier for each pixel.
Uganda For the purposes of this study, we shall rst examine the areas of
Uganda. Regarding the region of Katwe Lake, according to Census data obtained
from the Uganda Bureau of Statistics[
We proceed to examine the classi cation result extracted by the methodology that was described in the previous section.
We shall proceed to sum the pixels of the living areas (rgb(255,255,255)) in the region; The above process gives us a result of 2,242 total pixels that represent living areas in the compound of Lake Katwe. Thus, we calculate that the density multiplier of each pixel that represents living area in the compounds of Lusaka, shall equal to 11:5834.
Lusaka The same process is followed for the regions of Lusaka, where the
district of Kanyama is examined. According to Census data from 2010[
We then proceed to examine the classi cation results for Lusaka; it is noted that the Kanyama compound consists of three polygon shapes: (i) Old Kanyama (ZMB TW0029), (ii) New Kanyama (ZMB TW0050) and (iii) Kanyama West (ZMB TW0078). The aggregated result for those three subregions can be examined.
We shall proceed to sum the pixels of the living areas (rgb(255,255,255)) in the three districts using the same methodology. It is estimated that a result of total 64,199 pixels represents living areas in the compound of Kanyama. Therefore we calculate that the density multiplier of each pixel that represents living area in our Lusaka study area shall equal to 7:02352. After calculating the weight variables for both regions as described above, the clipped pixelated images along with the multipliers can be used as input to calculate the population of each region. For each compound, each white pixel is multiplied by the estimated weight variable of the respective study area (11.5834 for Uganda - 7.02352 for Lusaka). The outputs provide with the nal dataset containing the population estimations for all the subregions of each country.
It is important to note that the accuracy of the procedure described, depends largely on the classi cation output. A well classi ed image similar to the one obtained from the regions of Uganda during this study, can improve the results whereas a poorly classi ed region similar to the one obtained from Lusaka shall result to a dataset that di ers greatly from its actual population.
The results obtained from the described methodology shall provide us with the nal dataset containing information on the population data for each one of the study areas. However, the same procedure can be followed for estimating the dwellings of the regions of interest by appropriately parameterizing the census data input along with the forecasting model described in the previous sections.