A MULTI-SCALAR PHOTOGRAMMETRIC RECORDING APPROACH IN TERMEZ (UZBEKISTAN)

(1)

__________________________

* Corresponding author

A MULTI-SCALAR PHOTOGRAMMETRIC RECORDING APPROACH IN TERMEZ (UZBEKISTAN)

Jorge Angás^1*, Paula Uribe², Enrique Ariño², Josep M. Gurt³, Verònica Martínez-Ferreras³, Shakir Pidaev⁴

1 Universidad Politécnica de Madrid, Spain – [email protected]

2 Universidad de Salamanca – (uribe, argil)@usal.es

3 Universidad de Barcelona – (vmartinez, jmgurt)@ub.edu

4 Institute of Fine Arts, Academy of Sciences of Uzbekistan, Uzbekistan - [email protected] Commission II, WG II/8

KEY WORDS: Uzbekistan, Ancient Termez, CORONA Satellite, photogrammetric techniques, 3D reconstruction, open-source libraries

ABSTRACT:

This paper presents the preliminary results obtained during the 3D recording campaign carried out in 2018 by the Spanish-Uzbek IPAEB mission in the archaeological site of Termez (southern border of Uzbekistan). Ancient Termez is an important historical city within the Silk Road located in the ancient Bactria region. The archaeological work performed at the site since the beginning of the 20^th century allowed a large fortified urban complex to be identified that includes other walled enclosures inside it, i.e., a Hellenistic- Seleucid fortress founded after the campaigns of Alexander the Great in the late 4^th century BC, several Buddhist monastic complexes dated to the Kushan period (1^st to mid-3^rd centuries), and a large urban settlement dated to the Islamic period which includes the city proper or shahristan and the suburbs or rabad. After the destruction by Genghis Khan in 1220, Termez was rebuilt following a different plan. Major changes involved the movement of the pottery workshops from the rabad to the previous shahristan. The research focuses on: a) the identification, study and archaeological contextualization of ceramic production centres located in different areas of the ancient Termez from the Kushan to the Islamic period (1st to 14th centuries AD); b) the integration of the pottery workshops into the general topography of the site and c) the study of their evolution in relation to the transformation of the urban design. Since the site is currently located in a military area – close to the border area between Uzbekistan and Afghanistan –, the archaeological work is restricted to specific zones and the use of aerial devices such as drones is forbidden. However, this research requires both micro and macro spatial approaches to accurately record all the archaeological structures and to evaluate the integration and evolution of the pottery workshops into the general topography of the city. In order to fill this gap, declassified images of the CORONA satellite program were analyzed and compared to historical and archaeological data. In addition, we propose a geometrical and graphical recording and distribution system of the kilns –located in the rabad and the shahristan –and the ceramics produced and used in Termez during the period studied by means of photogrammetric techniques. The results are aimed at management through open-source 3D formats and web mapping GIS libraries combined with historical satellite information that defines the different archaeological areas.

1. INTRODUCTION OF THE ARCHAEOLOGICAL SITE: TERMEZ (UZBEKISTAN)

Termez is located on the right bank of the Amu Darya and near the mouth of its tributary, the Surkhan Darya. The region of Termez is characterized by an almost flat topography (mean altitude 300 m) and a few small hills that are the southernmost residual reliefs of the Baysuntau-Kugitangtau range. The Amu Darya-Surkhan Darya floodplain forms part of the Afghan-Tajik depression which comprises a Hercynian basement formed by highly metamorphosed Precambrian-Palaeozoic rocks and by Permian-Triassic sedimentary and volcanic rocks. The city was one of the most important urban centres in northern Bactria/Tokharistan, a historical region that included the present-day territories of southern Uzbekistan and Tajikistan (Figure 1), and northern Afghanistan. The exact date of the founding of Termez is unknown. It has been related to the campaign to conquer the ancient Bactria and Sogdiana by Alexander the Great (between 329 and 327 BC) as well as dated in the period of the first Seleucid kings (Leriche 2001; Pidaev 2001; Leriche/Pidaev 2007; Leriche/Pidaev 2008).

Termez includes several walled enclosures inside it. The earliest of them has provided evidence of Greek occupation. Thus, pottery in Hellenistic tradition has been recovered in the lower levels of the fortress (Citadel) located on the right bank of the Amu Darya. According to their features, the vessels should be dated to the late 4th or the early 3rd centuries AD (Pidaev, 1991). The second enclosure (Tchingiz Tepe) is located North- West of the Citadel. A recent archaeological investigation carried out in several sectors by our team, comprising a complete study of the archaeological sequence supported by radiocarbon analysis, allowed to date the construction of the walls between the early 2nd and the mid-1st centuries BC.

Therefore, the walls of Tchingiz Tepe could be dated immediately after the arrival to the region of the nomadic tribes known by the name of Yuezhi (Martínez et al. 2014; Gurt et al.

2015). Under the rule of Kushan kingdom (1st century AD – mid-3rd century AD) a new fortified area, the sharistan, was built. During this period, Termez became an important religious centre that hosted a number of Buddhist monasteries such as Fayaz Tepe, Kara Tepe, Tchingiz Tepe and Zurmala (Abdullaev, 2013; Stavisky, 2001).

The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W15, 2019 27th CIPA International Symposium “Documenting the past for a better future”, 1–5 September 2019, Ávila, Spain

(2)

In the mid-3r vassal state o complexes we In the 5th c established the the Buddhist reached its gr destruction of city witnessed the movement previous urban after this perio the 17th or 1 Barthold 2002

Figure 1. Loc close to the b

(Natural Ear (above). M

2. RESEA

The Spanish- Termez since documentation campaign (201 One of the rea archaeological mud and it (O’Grady et relevance to g

rd century AD of the Sasanian ere reformed, ch century AD a

emselves as aut monasteries w reatest extensio f the city carried d a period of u t of the pottery n area or shahri od, which ende 8th centuries ( 2, 58).

cation of the arc border area betw

rth, Google Sat Map of the ancie (Pida ARCH OBJEC

APP -Uzbek IPAEB e 2006, and n with differen 18).

asons for this i l structures of T

is difficult to al., 2018).

graphic and ge

D, the Kushan n Empire; the hanging their or succession o tonomous entiti were finally ab on in the Islam

d out by Gengh urbanistic chan y workshops fro

istan. Little is k ed in complete d (Karimov 2001

chaeological sit ween Uzbekista tellite view and nt settlement of aev, 2001b).

CTIVES: A M PROACH B mission has d a process nt scopes was in

s related to the Termez. Many o preserve the This phenome eometry record

kingdom beca Buddhist mon riginal function of nomadic pe ies in the regio bandoned. The mic period. Afte his Khan in 122 nges which inv om the rabad t known about Te depopulation ar

; Leriche 2001

e of ancient Ter an and Afghani

ArcGIS Image f Termez (below

ULTI-SCALA

s been workin of 3D geom nitiated in the e conservation o

of them are bu em once exca enon gives sp ding techniques

ame a nastic nality.

eoples on and e city er the 0, the volved to the ermez round 1, 81;

rmez stan ery)

w)

AR

ng in metric latest of the uilt of avated pecial s, not

only with the f The recor (Figu archa data more et al In th follo - The the S - Th the in - Th platf -To p the a archa The the Elev a DE allow chan set a archa today The with docu areas obtai a dig to pr the docu The form us to chan to t docu to 3 rema foun

as a scientific a view to any future.

data recorded rd of archaeolo ure 2), rangin aeological land

management a e important role l., 2019) combi his regard, all owing aspects as

e reconstruction Soviet-Afghan W

e geometric do nterests of its c he integration o form.

present differen archaeological s aeological struc first level deals archaeological vation Model (D EM previous to ws a compariso nges in the arch

allows us to co aeological stru y.

second level d special attent umented five ce s of the site. Th in a volumetric gital reconstruct reserve them on degradation s umented on a ye

third and last l ms. The photogr o reconstruct p nges, and obser

their volumetr umentation leve 3D record by ains as well at t nd.

guarantee, but kind of restora in Termez w ogical artefacts ng from tiny dscapes, althoug

and visualizatio e in the context ining dissemina the documen s a main objecti n of the archaeo War.

ocumentation of onservation.

of all the data nt hypotheses in sites and the po ctures.

s with the topog site with the DEM). The aim

o the Soviet-A on to the curren

haeological rem ompare, interpr uctures that are

deals with the tion to cerami eramic kilns dis he final objectiv

and spatial dist tion and a typol nce they have be uffered by th early basis.

level aims to d ammetry of the profiles, descri ve their degree ric evolution.

els mentioned ab photogramme the archaeologi

t also as an ind ation or consoli will be used in related to ceram ceramic sha gh the concept on is playing a xt of Cultural H ation and scien ntation acquired

tive:

ological landsc f each differen a on a 2D/3D n relation to the otential identific graphy and sate e construction

of the first leve Afghan War. Th nt DEM in ord mains. Furtherm ret and georefe

e not visible o recording of e ic kilns. In th stributed across ve of this docu tribution to be logy of the kiln een excavated;

hese materials develop a typolo

ese movable obj ribe and quant e of standardiza

In addition above, we took t

etric technique ical sites where

dispensable tool idation work in n a multi-scale mic production ards to whole t of multi-scale an increasingly Heritage (Boutsi ntific purposes.

d has had the ape previous to nt work level in web mapping e topography of cation of buried ellite images of n of a Digital el was to obtain his information der to ascertain more, this data erence different on the surface excavated areas his regard, we s three different umentation is to able to perform ns. It is difficult in view of this, needs to be ogy of ceramic jects will allow tify the formal ation according to the three the opportunity es some bone e the kilns were l n e n e e y i . e o n g f d f l n n n a t e s e t o m t , e c w l g e y e e The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W15, 2019

27th CIPA International Symposium “Documenting the past for a better future”, 1–5 September 2019, Ávila, Spain

(3)

Figure 2. Mu resolution a c

3.1 Archaeol Due to the imp the CORONA process in 199 CORONA sa States between provide great interesting ele obtained fro improvement.

include additio to the satellite Moreover, cam by the Nat (information p September 20 photogrammet processing co imagery is on images and c different reaso where morph capture of thi other natural o change the inte The acquisitio fitting them th dates taken by reduced to som These structur are easily ide after the War–

The advantag reconstructing changed the m These images comparison t evaluation of anomalies on t 3.1.1. Data ac First, the DEM using two ster 3), taken duri reconnaissance images were t board the KH images allow photogrammet be described stereoscopic i wide, the area 3),which mean be minimal (W camera with tw stereoscopic backwards, an satellite. The i

ulti-scale classif and size accordi carried out with

3. MET

logical landsca possibility of u A program were

95 of more tha atellite reconna n the 1960s and t potential ma ement detection om the con However, the onal informatio e position at t mera calibration

ional Geospat provided by th 018). Because

tric or a visu ontext is very ne of the best a

cartographic re ons. This is es hological chang

is information or anthropic eff erpretation of th on of these ima hrough common y a GNSS senso me architectural

res have remain entifiable in bo – .

ge of using the g the DEM befo morphology of th s have been th

o current sate f these differe

the ground.

cquisition M reconstructio reoscopic image

ing mission 10 e program, O taken on Octob H-4A satellite.

ws for creatin try software wi

later on. Th images is appr of study being ns that image d Watanabe et al.

wo lenses 609, images 30º fo nd a 70º total s images had a 17

fication of the g ing to Cultural H h different techn

THODOLOGY

ape. Use of sate using drones, th

e investigated.

an 800,000 ima aissance progr d 1970s during aterial for terri n thanks to th ntinuous sate e declassificati on related to th the instant of i n reports have n tial-Intelligence he USGS ERO of that, using ual analysis b difficult. Neve alternatives whe sources in spe specially impor ges have take due to the Sov fects. This facto

he archaeologic ages by the US n GCP support or. The control l structures and ned intact for a oth historical p ese images wa ore the war, wh he archaeologic he basis for ge ellite images.

ences is usefu

on of the 1960s es acquired from 012-1 of the A

ctober 17-20, ber 20, 1964) w

. The stereosc ng digital terr ith a series of li he surface co roximately 300 g located in the

distortion on th ., 2017). The J 602 mm in foc forward, stereo canning angle 7x232 km nomi

geometric recor Heritage typolo niques.

Y

ellite image he satellite imag

The declassific ages taken unde ram by the U the Cold War.

itorial analysis he spatial resol ellite perform

on process did he cameras use

image capture not been declas e Agency (N OS User Servic

these images ased on its d ertheless, COR en there is a la ecific places d rtant in those p en place since viet-Afghan W or could signific cal landscape.

SGS has allowe points between l points are bas d irrigation cana at least 60 year periods –before as the possibili hich is likely to

cal site.

eometric and v The detection ul to verify v

site was carrie m the USGS (F American COR 1964 (the sel with a J-1 came copic nature o rain models mitations which overed by the

0km long by 2 e central area (F he area of study J-1 was a panor cal width which oscopic images perpendicular t inal coverage.

rd by ogy

ges of cation er the United They s and lution mance d not ed nor time.

ssified NGA) ces in in a digital RONA ack of due to places e the War or cantly ed for n both sically als.

rs and e and ity of o have visual n and visual

ed out Figure RONA lected era on of the using h will e two 20km Figure y will ramic h took s 30º to the

In a auxil by th latter restit 2006

Figu DS

Seco phot DZB resol imag COR from are 9 and h Third glob web on w mate

Ta

addition to the liary cameras t he individual l r provided ex tution of the i 6).

ure 3. Stereosco S1012-1039DF

orange) on the ond, the rest of o-interpretation B00403800056H

lution in order ge was taken RONA reconna m July 1963 to J 9 inches wide, v have a resolutio d, all the imag al coordinate s Mercator) with web applicatio erials and result

able 1. Classific that

J-1 camera, t that took image lenses, of the h ternal orientat images (Galiat

opic images use 157 (above) and ground (below f the images o n process. Am H011001 provi

r to identify a during missi aissance progr June 1967. The vary in length fr on of 0,60 to 1,2 es and data we system, in this h coordinates e ons, which wo s in a web ma

cation of the CO have been used

the satellite h es of the total horizon, and o tion in the ph tsatos, 2004; G

ed: DS1012-10 nd their footprin w) (USGS-Earth on Table 1 we mong all of ided us with th archaeological ion 4038 of ram, which w e Keyhole (KH from 4 inches to

21 m. (2 to 4 fe ere integrated u s case EPSG 3 expressed in m ould allow us apping environm

ORONA images d for this study.

ad a series of space captured of the sky. The hotogrammetric Goosens et al.,

39DA163 and nts (green and hExplorer).

ere used in the them, image he best spatial remains. This the American was operational ) KH-7 images o 500 feet long, eet).

under the same 3857 (WGS 84 meters and used to use these ment.

s declassified f d e c ,

e e l s n l s , e 4 d e The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W15, 2019

(4)

3.1.2. Image processing

The photogrammetric process includes a total of 52 control points in order to achieve better adjustment in the spatial referencing of the block of images during the photogrammetric restitution process. These points were taken in the field in April, 2018 using a GNSS sensor with sub-metric accuracy. Their distribution is related to the localization of a series of interesting archaeological elements which were worth documenting during this fieldwork campaign. These GCP occupied a small space within the whole image. Therefore, we used QGIS-digitizing tools to digitize a series of points on the georeferenced CORONA images. Although the accuracy of the resulting coordinates may be questionable, they were very easily localized on the images during the photogrammetric process.

This process is justified in view of the lack of a wide orthophotography of the area and of the impossibility of creating it ourselves as drones are not allowed in this zone.

We mainly used two software products for photogrammetric restitution. One is ERDAS Imagine Photogrammetry; in the non-metric cameras category we worked with focus width (609,602 mm) and the resolution used in scanning (7 µm) to set the interior orientation. This model is fit to use with the CORONA images free from interior calibration parameters of the main and fiducial points (Altmaier and Kani, 2002;

Galiatsatos, 2004). We also used The Automatic Terrain Extraction tool (ATE), which allows for obtaining DEM from the restituted photogrammetric block (Bellido, 2018).

Next, we performed a photogrammetric restitution based on the SfM technique in Agisoft Metashape Professional (Watanabe et al., 2017). Like Imagine Photogrammetry, interior orientation was obtained based on focus width and scanning resolution.

Fiducial points cannot be localized given that they are inexistent, thus distortion rates are left blank so that the software can calculate them during bundle block adjustment.

Following the bibliography on CORONA image correction (Shin and Schenk, 2008) and on rotation angle measurement based on GPS data (Wierzbicki and Krasuski, 2015), we carried out a first estimation of the exterior orientation parameters (Table 2):

Image DS1012-1039DA163 DS1012-1039DF157 Xo 7.477.170 m 7.489.300 m

Yo 4.485.140 m 4.490.290 m

Zo 185.000 m 185.000 m

Yaw 244º 244º

Pitch -16, 375º -16, 375º

Roll 0º 0º

Table 2. Table of the initial exterior orientation parameters of the images.

Both photogrammetric restitution processes are aimed at obtaining an orthorectified and georeferenced image. In this regard, numerous approximations have been performed by changing the images used in multiple combinations: whole images, images cut over the area of study, non-georeferenced images, images georeferenced with 2^nd degree polynomials and images georeferenced with 1^st degree polynomials (affine transformation). Furthermore, both GPS and QGIS-digitized control points have been interchangeably used to obtain the desired result.

Unfortunately, we were not able to perform the DEM with the selected CORONA images due to overlapping problems between them and a lack of key data, unclassified by the NGA for photogrammetric restitution. Consequently, in order to reconstruct the DEM of recent times, we obtained altimetry data from the relief map by Dumont, Reynard and Person in the context of the French-Uzbek archaeological mission in Bactria

(Pidaev, 2001a). Figure 4 shows the altitudinal data in level curves at 0,5 meter intervals and in a series of bench marks and topographic points. The localized walls and buildings of the site appear as auxiliary information providing context to the relief.

Figure 4. Interpolated 3D model of current period with the location of the ceramic kilns (above). Longitudinal section on the DEM between the T5 and T8 area of ceramic kilns (below).

As to archaeological interpretation, the image has been adjusted in different ways in order to improve the visualization of remains still buried. First, we performed basic contrast, brightness and histogram adjustments. Next, the work method’s algorithms were applied to the LiDAR data, namely Analytical Hillshading, Hillshading from multiple directions, Simple Local Relieve Model, Principal Components Analysis, Sky-view Factor and Positive/negative openness.

3.2 Ceramic kilns and archaeological excavations.

Terrestrial photogrammetry 3.2.1. Data acquisition

The second documentation level involved acting over different ceramic kilns in the site. Due to logistic difficulties in the field, we used those photogrammetric techniques that would allow for rapid extensive documentation.

Regarding the photogrammetric system used for the data extraction of the kilns already excavated, a 5 meter pole with a micro four thirds Olympus EPM 1 camera and a Zuiko Digital ED lens 12mm f2 perpendicular to the ground was set up. The final height obtained with this system was 5,7 meters with an angle between 40° and 45° in order to facilitate data collection with a linear movement every two meters, both longitudinally and transversal to the excavation so as to obtain a redundant overlap between images that was valid for three-dimensional reconstruction.

(5)

Figure 5. Data the ceramic

Taking into dimensions of the resolution mm), we obta with a resolut between long (Figure 5). Th archaeological approximately resolution, wh a better resol work carried o excavation of ground due to well as the la with a laser di

Figure 6. Isom

a acquisition us kiln T2 H2 and

account the f the sensor used

of the camera ained an 8,2 m tion of 2 milli gitudinal and tr

his technique w l site as it y 2000 square hich guaranteed lution obtained out in T8 H1 al f this area, base

the humidity o arge ash stains.

stance measurin

metric view on PDF 3D of the

sing photogramm d results (orthop

height of the d (micro four th (12 Mpx.) and meters x 6,2 me imetres per pix ransverse imag was used at se allows digit meters in jus documentation d in the final o

llowed us to fin ed on the chan of the mud/sun- The metric re ng device.

the 3D section e ceramic kiln T

metric techniqu photo and sectio

e camera and hirds 17.3 x 13 the focal lengt eters image foo xel, with an ov ges of 65% - everal points o tizing an are t one hour in n in a short time

orthophoto set nd a kiln prior t nge of colour o -dried brick wal ferences were

generated in fo T5 H1.

ues in on).

d the mm), th (12 otprint verlap 70%

of the ea of high e with . The to the of the lls, as taken

ormat

3.2.2 The Meth form woul creat brow

3.3 3.3.1 For phot Term hom full Chec this is ai main geom us to the perio In sp ceram etc), locat

F

3.3.2 This Profe spec

2. Image proce documentation hashape Profess mats with a vie ld be easy to m ted a web plat wser in high reso

- 3D point based po Potree ba - JSON me

for inform - Web map sites usin (Figure 1 Ceramics. Clo 1. Data acquisi

the final do ogrammetry tec mez laboratory

ogeneous light frame camera cker calibrated geometric reco imed at genera n pieces reco metric and grap o extract and to ceramic forms od.

pite of the lon mic remains ta this technique tions with comp

Figure 7. Data a portable clos

2. Image proce data was also fessional. In this ific way given

ssing n collected w sional. The resu ew to facilitati anage by the w tform that coul olution and diff clouds on Potr int cloud rend sed upon three eshes for the b mation sharing (

pping of each g Leaflet and o 1).

ose Range Phot tion

cumentation l chniques on the

close to the s system, a rota with a fixed colour chart (F onstruction proc

ating a three-d overed that al hic millimetre chronologically s and their di

ng hours that akes (rotations e allows for cru

plicated logistic

acquisition ceram se range photog

ssing o post-processe s case, each pie that we applied

was processed ults were expor ting a typologi whole technical

ld be accessed ferent types of s ree (free open- derer for large e.js library (Sch browser viewer (Figure 6).

orthophoto of open-source Jav

otogrammetry

level we used e main ceramic site. We desig ating plate, wit 50 mm lens Figure 7). The cess with low c dimensional rep llows for rep details. This fa y divide the sta imensions in e

post-processin s, masks, align ucial quick dat cs.

amic process by grammetric tech

ed with Agiso ece was proces d a minimum of

using Agisoft rted in different ical study that team. Thus, we d online with a

scale:

source WebGL point clouds).

huetz, 2016).

r and 3D PDF the excavation vaScript library

d close range c vessels in the ned a portable th a Canon 6D and a Colour automation of cost techniques pository of the producing any actor will allow andardization of each historical ng each of the nment, scaling, ta collection in

y means of a hniques.

oft Methashape sed in a certain f four complete t t t e a

L . F n y

e e e D r f s e y w f l e , n

e n e The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W15, 2019

(6)

turns per piece lighting, takin was scaled, th Potree convert Furthermore, t for use on a 1.9.1.js, Hand

Figure 8. 3D recorded in

First, the resu objective of su system that co a) Three-dime The ceramics, were converte WebGL Potre 8).

Figure 9.

DZB00403800 contrast; c) assessment [p bit); d) Princip

16 directio b) Orthophoto satellite imag

e at different h ng each photo e he results were rter so as to fin the final meshe

web browser dpas_1.0.1.min.

D reconstruction Termez and the PDF and 4. R ults obtained w ubsequently gen ombines:

ensional models kilns and archa ed to point clo ee. Parallel 3D

Different visua 0056H011001 ; ) Simple Local R pixels]: 20, histo

pal component ons of Hillshadin

os of the kiln ges. Orthopho

eights with the every six degre

converted to p nally use this p es were export with HP3D_0 .js, and three.j

n of the differen eir geometric us JSON formats RESULTS were analyzed nerating an inte s (point clouds, aeological exca ouds through th PDF files wer

alizations: a) CO

; b) The same im Relieve Model ogram equal wi analysis of Hil ng and sun elev ns, excavated a otos were imp

same homogen ees. Once the m point clouds wit

latform as a vi ted to JSON m 0.0.4.min.js, jq

s libraries.

nt ceramic obje se by means of .

separately wit egrated web map

meshes).

avations docum he free open s re generated (F

ORONA Image mage with incre

(Radius for tren ith 2% cut-off f llshading (Band vation angle 35º areas and hist plemented in

neous model th the iewer.

models query-

ects f 3D

th the apping

mented source Figure

e eased end

for 8- d 1 of

º).

torical web

mapp to ge and c The failu dispo thank of po the p As t (Figu COR been (DZB impr best Mod morp (Kok

ping using the eospatially rela current orthoph reconstruction ure to obtain an osal. However, ks to the DEM oints with altitu previously cited o the archaeolo ure 9, 10, 11) RONA image w n georeference

B00403800056 rovement in ord results were del. This algorit phological elem kalj and Hesse,

open source Le te different lay hotos.

of the site coul old DEM from we were able M (Figure 4) gen

udinal data extr d relief map of t ogical interpret has been carr with the highes ed and tiled

H011001). In der to identify th obtained with thm is based on ments to revea

2017).

eaflet library. T yers of CORON ld not be compl m the CORONA

to reconstruct nerated from th racted from the the site in the a tation, the phot ried out based

st resolution, w d for the L

n terms of he archaeologic h the Simple n eliminating th

al the smalles

The objective is NA information leted due to the A images at our the current site he interpolation level curves in ancient Termez.

tointerpretation mainly on the which has also Leaflet library f the visual cal remains, the Local Relieve he most notable st scale record s n e r e n n . n e o y l e e e d The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W15, 2019

(7)

Figure 10. T CORONA Ima

a) CORON

Archaeologica image. c) Goo 5. CONCL The study de represents a k many cases it the pre-war ar have provided the site, but w archaeological and the begin mechanization sites studied allows for com archaeological spectral amp archaeological Its use allows of the archae information w types of 2D or the different s layers which converting to JavaScript lib clouds and JSO In spite of th difficulties in adjusted spatia to complete th difficulty deri was excluded methodology requirements i study.

Even when it restitution p georeferencing elements of in the visual dia requires polyn a lack of carto by the time dif of control poin

Figure 11. F view, dia interpretation

The archaeolo age DZB00403 NA image

al interpretation ogle Satellite vi

LUSION AND emonstrates th key source of in

is virtually the rchaeological la d useful updated would not have l landscape 50 nning of the 1 n and urban de at that time w mparing groun l structures. H plitude does n

l structures of s us to obtain a eological lands with web mappi r 3D informatio satellite images

allows for ove o GeoJSON brary. As to 3

ON meshes mo he great potent nvolved in cor al information m he DEM recon ives from a lac

from the decl applied here in order to add is not possible rocess, the g is an altern nterest or using chronic analysi nomial geometr ographic referen

fference betwee nts.

Final web mapp achronic images

n and georefere typology (ce

ogical interpre 3800056H01100

DZB00403800 n from DZB00 iew.

D FUTURE DE at the use of nformation in a e only one for t andscape. The d information on e allowed the r

years ago. Th 1970s period p evelopment and were well prese nd anomalies to owever, the sp not allow a small dimension

joint reading o cape if we co ing systems tha on. 2D data wa s and interpretin

rlapping two im (Figure 11) D models, we ostly in the three tial of the CO rrecting their g

means that it h nstruction of an

ck of auxiliary lassification pr includes a dress this short e to carry out t digital proce native when it g the images as is of the territo ric correction w nce resources i en the localizati

ing that include s of CORONA, enced links to d eramics and kiln

tation through 01.

0056H011001.

403800056H01

EVELOPMEN f CORONA im

archaeology sin the reconstructi

use of drones w n the current st reconstruction o he end of the 1 precedes agricu d thus many o erved. Above a o corroborate b patial resolution detailed stud ns.

of the reconstru ombine this typ

at integrate diff s organized by ng them in diff mages, drawing using the L e used Potree

e.js library.

ORONA images geometry to o has not been po

ncient Termez.

y information w rocess of 1995

series of aux tcoming in this the photogramm essing and i

t comes to fin s reference poin ory. This proce which is hindere

in certain place ion and the che

es: Google satel archaeological different 3D mo

ns).

h the b) 11001

NTS mages nce in ion of would tate of of the 1960s ultural of the all, it buried n and dy of uction pe of fferent tiling fferent g and Leaflet

point s, the obtain ssible . This which . The xiliary s case metric image nding nts in essing red by es and ecking

llite l dels

As decla issue help know expo the p adva auxil to id with orien In th camp the C altitu of m Fina colla HTM whic infor exce of th onlin morp easy addit be p deve

This (HAR and all fu Univ camp this p Caja

Abdu in ol Altm gene Phot https Belli prog arqu Fin d Bout onlin Phot 173- 173- Galia Reso COR

to the NGA assification pro e that should b

the scientific c wledge on th onentially espec processes asso anced in spite liary informatio dentify interesti

the new tec ntation and the he next few y paigns with the CORONA satel ude photogramm multispectral sen

lly, the conclus aborative deve ML5 Canvas is ch favour the rmation associ eptional multi-s he requirements ne and offlin phologic inform

to use with tional software programmed in elop in the futur AC work was su R2016-75133-C

CONCERAC ( unded by the S versities. Found

paign in Terme paper through t al (RYC-2014-1

ullaev, K., 201 d recent finds. P maier, A., and

eration from CO togrammetry a s://doi.org/10.10 ido, A., 2018.

grama Corona eológico del y de Máster. Univ tsi, A.-M., Ioa ne visualization togramm. Remo 180. https:/

2019

atsatos, N., Don olution Elevat RONA Imager

failing to inc ocess of the C e debated as th community in t his images an

cially in the fie ociated with ph of being clo on used throug ing elements is chnological too exterior orienta years we will purpose of ver lite in the field metric techniqu nsors.

ion in terms of lopment platfor that this allows e use of thes ated with a d cale documenta s we consider e ne measuremen mation extractio

a browser s , and above all, nto any implem re.

CKNOWLED pported by tw C3-1-P) led by (HAR2016-751 Spanish Ministr dation Palarq ez in the year 2

the post-doctor 5789).

REFEREN 3. The Buddhis Parthica, 15, pp Kany, C., 2 ORONA satelli and Remote S 016/S0924-271

Análisis del p a través de la acimiento de T versidad de Zar

nnidis, C., and n of complex ote Sens. Spati //doi.org/10.519

noghue, D. N. M tion Data D ry with Mini

clude auxiliary CORONA imag his information the use of these nd their use eld of remote hotogrammetry osely related.

ghout the COR s not likely to ols for a cam ation of images

continue to c rifying the data d, and we will a ues described w f using JavaScri

rms (GitHub) s for a vast set se models w database. They tation tool whic essential for our ent and graph on, use of filte supporting We , it is an open s mentation we

DGEMENTS wo R&D&I pro

y V. Martínez 33-C3-3-P) led ry of Science,

supported the 2018. V. Martí ral research con

NCES st culture of A p. 157–187.

2002. Digital ite images. ISP Sensing, 564,

60200046-1 potencial de las

la reconstrucció Termez (Uzbek ragoza.

d Soile, S., 20 x 3D geometri tial Inf. Sci., X 94/isprs-archiv

M., and Philip, Derived from imal Ground

y data in the ges, this is an n would clearly e images. First, would grow sensing, where y are not very However, the RONA program be compatible mera’s interior

in space.

carry out new a obtained from also use the low with other types ipt libraries on

together with of possibilities with contextual y constitute an ch meets many r research: high hic capability, ers and texture, ebGL with no system that can might need to

ojects, CERAC and J.M. Gurt, d by E. Ariño, Innovation and archaeological nez worked on ntract Ramón y

Ancient Termez

surface model PRS Journal of pp. 221-235.

s imágenes del ón del paisaje kistán). Trabajo

019. Interactive ies, Int. Arch.

XLII-2/W9, pp.

es-XLII-2-W9-

G., 2008. High Stereoscopic Control: An e n y , w e y e m e r w m w s

h s l n y h , , o n o

C , , d l n y

l f .

l e o

e . . -

h c n The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-2/W15, 2019

(8)

Approach Using Ikonos and SRTM Data. Photogrammetric engineering, 14.

Galiatsatos, N., 2004. Assessment of the CORONA series of

satellite imagery for landscape

archaeology: a case study from the Orontes valley, Syria, Durham theses, Durham University. Available at Durham E- Theses Online: http://etheses.dur.ac.uk/281/

Goossens, R., De Wulf, A., Bourgeois, J., Gheyle, W., and Willems, T., 2006. Satellite imagery and archaeology: the example of CORONA in the Altai Mountains. Journal of Archaeological Science, 336, pp. 745-755.

https://doi.org/10.1016/j.jas.2005.10.010

Gurt Esparraguera, J. M., Ariño Gil, E., Martínez Ferreras, V., and Pidaev, S. R., 2015. The Buddhist occupation of Tchingiz Tepe Termez, Uzbekistan in the Kushan period through the ceramic contexts. Archaeological Research in Asia, 3, pp. 19- 33.https://doi.org/10.1016/j.ara.2015.04.003

Hamandawana, H., Eckardt, F., and Ringrose, S., 2007.

Proposed methodology for georeferencing and mosaicking Corona photographs. International Journal of Remote Sensing, 281, pp. 5-22. https://doi.org/10.1080/01431160500104400 Karimov, I.A., 2001. Termez – an Ancient and Modern City at an Important Crossroads. Tashkent, Publishing-poligraphic share company ‘Sharq’.

Kokalj, Žiga and Hesse, Ralf, 2017. Airborne Laser Scanning Raster Data Visualization. A Guide to Good Practice. Ljubljana:

Založba ZRC.

Leriche, P., 2001. Termez antique et médiévale. In P. Leriche, S.R. Pidaev, M. Gelin, K. Abdullaev and F. Fourniau ed., La Bactriane au carrefour des routes et des civilisations de l’Asie centrale, pp. 75–99. Maisonneuve and Larose, Paris.

Leriche and Pidaev, S.R., 2007. Termez in Antiquity. In J.

Cribb and G. Herrmann ed., After Alexander. Central Asia before Islam, pp. 179–211. Proceedings of the British Academy 133. University Press, Oxford.

Leriche, P. and Pidaev, S.R., 2008. Termez sur Oxus: Cité- capitale d'Asie Centrale. Maisonneuve and Larose, Paris.

Leriche P., Pidaev S. R., Annaev T., D'Alascio H., Gavison S., Genequand D., Houal J.-B., Moustafakoulov S., Naoumova E., Prisco A., 2002. Mission archéologique franco-ouzbèque de Bactriane septentrionale. Rapport 10, 2002, 78.

Martínez Ferreras, V., Ariño Gil, E., Gurt Esparraguera, J. M., and Pidaev, S., 2014. The Enclosure of Tchingiz-Tepe Ancient Termez, Uzbekistan during the Kushan and Kushan-Sassanian Periods. Iranica Antiqua, 49, pp. 413–469.

https://doi.org/10.2143/IA.49.0.3009247

O’Grady; Caitlin R., Luke, Christina, Mokrišová, Jana, and Roosevelt, Christopher H., 2018. Interdisciplinary approaches to understanding and preserving mudbrick architecture in regional and diachronic contexts. Cogent Arts and Humanities, 5, pp. 1- 25.https://doi.org/10.1080/23311983.2018.1553326

Pidaev, S.R., 1991. Kyeramika gryeko-baktriyskogo vryemyeni s gorodishcha starogo Termeza, Sovyetskaya Arkhyeologiya, 1.

Academy of Sciences of USSR, Moscow, pp. 210–224.

Pidaev, S.R., 2001a. Contribution à l’histoire ancienne de Termez. In P. Leriche, S.R. Pidaev, M. Gelin, K. Abdullaev, F.

Fourniau ed., La Bactriane au carrefour des routes et des civilisations de l’Asie centrale, pp. 47–57. Maisonneuve and Larose, Paris.

Pidaev, S.R, 2001b. Mosquées de quartier dans l'ancciene Tirmidh (Ouzbékistan). Archéologie islamique, pp.61-74.

Schmidt, M., Goossens, R., Menz, G., Altmaier, A., and Devriendt, D., 2001. The use of CORONA satellite images for generating a high-resolution digital elevation model. In IGARSS 2001. Scanning the Present and Resolving the Future.

Proceedings. IEEE 2001 International Geoscience and Remote Sensing Symposium Cat. No.01CH37217 Vol. 7, pp. 3123-3125.

Sydney, NSW, Australia: IEEE.

https://doi.org/10.1109/IGARSS.2001.978277

Schütz, M., 2016. Potree: Rendering Large Point Clouds in Web Browsers. Faculty of Informatics at the Vienna University of Technology.

Shin, S.-W., and Schenk, T., 2008. Rigorous Modeling of the First Generation of the Reconnaissance Satellite Imagery.

Korean Journal of Remote Sensing, 11.

Tsantini, E., Martínez Ferreras, V., Ariño Gil, E., Gurt i Esparraguera, J. M., and Pidaev, S. 2016. Pottery Production in the Buddhist Communities in Central Asia: The Kushan- Sassanian Pottery Workshop of Kara Tepe Termez, Uzbekistan:

Kushan-Sassanian pottery workshop of Kara Tepe Termez, Uzbekistan. Archaeometry, 581, pp. 35-56.

https://doi.org/10.1111/arcm.12161

Ur, J. 2003. CORONA satellite photography and ancient road networks: A northern Mesopotamian case study. Antiquity, 77295, pp. 102-115.

Watanabe, N., Nakamura, S., Liu, B., and Wang, N. 2017.

Utilization of Structure from Motion for processing CORONA satellite images: Application to mapping and interpretation of archaeological features in Liangzhu Culture, China.

Archaeological Research in Asia, 11, pp. 38-50.

https://doi.org/10.1016/j.ara.2017.06.001

Wierzbicki, D., Krasuski, K., and Rykach, S. P. W. 2015.

Estimation of rotation angles based on GPS data from a UX5 Platform. Measurement Automation Monitoring, 61 11, pp. 516- 520.