DOI:10.1109/DigitalHeritage.2015.7419505

Extracted Abstract:

With togrammetry tion of point cl archaeological vations, how e rs requires a d ially locate ar y 3D or aerial ion of our dig digital GIS ne designed fo ctly connects t rd sub-centim kers, loci boun line enables th pective maps f ackend it uses import variou hosted by on-li ication of Arch y Iron Age sit dan. x Terms—A alization, str rmation syst rding Over the past become com gration of dig K GPS), low c ware. The qu ome integrated dle what wi rdings, and d yah, Southern research and writ ullah University o Kin of S ne gital A ension the rapid a among the loud ‘data scaf sites is now ever, the conti digital toolset in rtifacts, install l scan. We pr ital field recor 3D Top Plan or visualizing to Total Statio meter measur ndaries, and c he generation from the first a PostGIS d a s vector, raste ine geo-refere hField C++ to te of Khirbat Archaeology, ructure-from-m ems (GIS), I. INT t decade, digi mmonplace ital surveyin g cost tablets, a uestion now d within the ill become a igital docume n Jordan, the ting was funded i of Science and Te Neil G. Sm ng Abdullah U Science And T eil.smith@kau Archae n: Arc adoption of la archaeologica ffolds’ and dig w becoming co inual exposure n which to rec lations, and l resent ArchFie rding software production w massive 3D d ons and our R rements for a camera positio of publishable day of excavat atabase and th r, DEM and 3D enced database our 2014 field t al-Jariyah lo cultural he motion, LiD Level-of-Deta TRODUCTION ital archaeolog and cost ef g equipment (T and especially is how do archaeologist s an overload entation. In 20 excavation in part by NSF IG echnology. mith University Technology ust.edu.sa ology chField Fie aser scanning al community gital document ommon. In e of archaeolo ord, categorize oci within a eld C++, the l e that enables within a rend datasets. Arch RTK GPS uni artifacts, scan ons. The proce e orthographic tion to the last he ability to ex D datasets tha es. We presen d excavations o ocated in Sout eritage, scie DAR, geogra ail, Digital gy field recor ffective with Total Stations y photogramm these instrum s’ core toolbo of daily s 014, at Khirba methodology GERT and King Univers m y Field d C++ eld Wo and y the tation field ogical e and site’s latest real- ering hField its to nning essing c and t. As xport at can nt the of the thern ntific aphic field rding the s and metric ments ox to cans, at al- was fun d (Stru mas gig a rend by a daily field not o mas C+ + evol been ongo softw and spr e betw data plan Fig. data s Matthew Ho sity of Califor mdh5169@gm d Reco + a 4D ork damentally ch ucture-from-M sive 3D data apixel resolut dered simultan any off-the-sh y deluge of d d it would nee only vector ar sive point clo + was develop lution of an n developed o oing field ex ware to comb LiDAR/SfM) eadsheets. Int ween the field a from the fie ns and queryab 1. ArchField C++ sets. owland rnia, San Dieg mail.com ording D GIS hanged when Motion) was asets of point tion orthopho neously and o helf software. data acquisitio ed to render a rtifact position ouds and 3D ped to address integrated fie over the past xcavations. bine high prec ) with superv tegrated data d excavations eld to be imm ble field report + main gui showi Univergo g in the for Di daily aerial a introduced. t clouds, text otos that cou ver different d In addition n [18] to be i a digital Top P ns and loci but meshes. In re s these proble eld recording five years in ArchField p cision spatial r visor’s observ abases are s and lab analy mediately visu ts in real-time ing Khirbat al-Jar Thomas E. rsity of Califor tlevy@ucs e 4th igital and terrestrial The result tured meshes, uld not be e days of excav n, in order for ideally used i Plan that cont t multiple laye esponse, Arch ems. It is the software th a coordination provides a un recordings (su vations and d seamlessly sy yses to enable ualized as 4D e. riyah (KAJ) exca . Levy rnia, San Dieg sd.edu SfM was , and easily vation r this in the ained ers of hField latest at has with nified urvey digital ynced e raw D top avation go 978-1-5090-0048-7/15/$31.00 ©2015 IEEE Authorized licensed use limited to: TU Wien Bibliothek. Downloaded on October 26,2024 at 11:09:51 UTC from IEEE Xplore. Restrictions apply. In reco exc a C++ arch the f Arc h reco and carr i 3 topic Perh arch syst e be st [17] map fro m (reco them arch co m com exca artifa surfa mult Fig Sit n this paper, rding in th e avation) and + are addresse haeological fie field of scienti  Time-lap textured  An intui annotati o  Multi-DB hField C++ re rding and vis integrated in ied out in vari 3D/4D GIS so c of research haps the firs aeologists is em where diff tored and conn and ETANA servers that c m traditional orded in Acce m on online. T aeological fie mputer vision mbines plan re avations and facts and featu aces and artif ti-view stereo. g. 2. Aerial Map es excavated or s the fundame e 4 th dimens current soluti ed. The softw eld work (fig ific field recor psed viewing meshes and o tive interface on of 3D datas BMS Integrati epresents a no sualization th nto any archa ious regions th II. LITERA oftware for arc and developm st implement DATARC H fferent media, nected to a rel A-GIS [9] wer could take ba excavation te ess© or other s The one softw eld acquisition project [8]. eports with c multi-view c ures. The futur facts in 3D s . of Southern Edo sounded by the EL ental challeng sion (time s ions integrate ware is a 4D . 1). Its majo rding include: of massive orthophotos. e for user edi sets ion: SQL + N ovel breakthro hat can be im aeological exc hroughout the ATURE REVIEW chaeology has ment for the p ation of a H [5], an im primarily dig lational databa re designed as asemaps (gene echniques) an spreadsheet pr ware specifica n is REVEAL This is a re continuous vi camera captu re goal of the space using te om research area, LRAP/L2HE proj ges in digital sequences in ed into ArchF D GIS tailored or contribution : 3D point cl o ting, drawing OSQL ough in 4D di mmediately app cavation curr world. W s been an ong past three dec tailored GI S mage manage gital photos, c ase. Archaeolo s open-source erated in ArcG nd database t rograms) and s ally developed L, an NSF fu ecording tool ideo recordin ures of impo project is to o echniques suc , depicting the sh jects (Map data: G field n the Field d for ns to ouds, g and igital plied ently going ades. S for ment could oGIS e GIS GIS© ables serve d for unded that ng of ortant orient ch as proj (e.g . GRA stud [11] O tool doc u GIS S large deve high envi cam into ass o R man and app r can pro p capa the m M Post of-D high view can I of th deve for inst r one harp elevation ch Google, Digital G In contrast to ects have ado . ArcGIS) or ASS, OSSIM dies in scholar . Others have s with LiD A ument excava . Several recen e datasets in eloped a VR- h poly count ironment. The mera for fly-thr the scene c ociated metada Recently, wor nage a 3D GIS other GIS rel roach is that a be immediate perly imported ability of Arc models and tex MayaArch3D tGIS server [1 Detail (LOD) her resolution wing windows stream large c In contrast to he few archa elopment and archaeologist ruments utiliz more evoluti hange between th Globe 2012, Land o these in hous opted off-the- ropen source M, QGIS) to d rly journals or focused prim AR or SfM (e ations but do n nt publication n a 3D GIS -GIS that opt geometric m eir system su roughs and sk can be mani ata. rk by [4] use S that include lated datasets. all the analytic ely applied to d. A major dr Scene requiri xtures importe utilizes We 1][3]. This pro models in its models by e s. A major adv content over th all of these p aeological fiel a full GIS DB s that does r zed in field ex ionary step be he lowlands and h dSat7) se software de -shelf proprie e GIS program digitize their r in online da marily on inte e.g. [12], [2], not integrate t ns have attem S environmen timizes the v models in a 3 upports path keletal animati ipulated or s ed Esri’s ArcS s textured mo . A significant cal and editing the archaeolo rawback is the ng the author ed into the sof ebGL and di ogram can han s navigable v effectively op vantage of this he internet to m projects, Arch ld recording BMS. It is the real-time reco xcavations. W eyond its prev highlands of Anc evelopments, m etary GIS soft ms (e.g. Map data and pu atabases [12], egrating surve [6], [7], [19 the results in mpted to inte nt. In 2010, visual renderin 3D archaeolo animation o ion. Objects lo selected to a Scene as a to odels, artifacts t advantage o g tools of ArcS gical datasets e limited rend rs to simplify ftware. irectly links ndle limited L iewer and ha ening indepen s software is t many users. hField remains tool with ong only tool desi ording with d ith ArchField vious version cient Iron Age E many ftware pInfo, ublish [20], eying 9]) to a 3D egrate [10] ng of ogical f the oaded access ool to s, loci of this Scene once dering both to a Level- andles ndent that it s one going igned digital dC++ s has Edom. Authorized licensed use limited to: TU Wien Bibliothek. Downloaded on October 26,2024 at 11:09:51 UTC from IEEE Xplore. Restrictions apply. been made by making 4D field recording and visualization possible. III. A 4D GIS FOR FIELD RECORDING In recent years, a number of conference topics have addressed the idea of a 4D GIS (c.f. DHIC 2013 and CAA 2015 abstracts). In certain respects, every digital excavation is four-dimensional and once imported into current 3D GIS tools such as ArcScene it can be considered a 4D GIS (e.g. [4]). However, these current solutions do not fully address the underlying problem: archaeology is an inherently destructive science necessitating continual advancement in field recording tools that achieve an increasing improvement in accuracy and comprehensive recording. 4-dimensional GIS field recording should be envisioned as the new objective in digital archaeology requiring a fundamental change in both methodology and software. In this light, four core objectives of a 4-dimensional field recording methodology can be proposed that will require an ushering in of a new era of digital field recording. The purpose is to address more fully the endeavor by archaeologists to conduct scientifically based excavations that properly examine all available archaeological evidence to test anthropological and historical theories about past culture. Objective 1: 3D Scanning on a daily/hourly basis A core objective is digital 3D scanning of the current excavation on a daily basis. Each day’s 3D scan captures the newly exposed layers so that over the entire period of excavation every sedimentary change is documented. On an hourly basis as a new artifact, installation or even collapsed wall is discovered in situ, it’s excavation and eventual removal is fully documented with 3D scanning. This can be accomplished using TOF (Time-of-flight) Laser scanning, but photogrammetry may prove the most efficient approach. Using the widely adopted photogrammetry software amongst the archaeological community, an entire and complete (minimal occlusion) excavation can be captured in minutes either from the ground or air [23]. The new problem becomes how to store, manage, co-register, and visualize such massive datasets being generated on a daily or hourly basis. Objective 2: Integration of 3D layers with a spatial DBMS (Digital Database Management System) In order to qualify as a 4D GIS, the 3D data must be integrated spatially with all other time-sensitive database entries. This would include other 3D measurements (e.g. GPS or Total Station), digital images, field databases, field notes, and later auxiliary tables generated from specialist analyses of artifacts, loci, stratigraphy or architecture. This integration would require 3D layers to be able to represent three different types of time: 1. Field Excavation Processes: The ability to represent the artifacts, loci, architecture, sedimentary layers exposed on a specific day of excavation (e.g. a 3D Top Plan). 2. Stratigraphy/Phasing: Represent an entire stratigraphic layer of a specific time of occupation by the original inhabitants. This would require the ability to represent the key 3D scanned architectural elements associated with only that period of occupation. 3. Site Abandonment and Deposition: Model the depositional layers produced through natural and man- made post-abandonment processes. This could include modeling a series of collapsed walls, how upper-stories overlay lower floors, the decay of organic materials, and even the movement/levitation of artifacts through erosional processes. Objective 3: Real-time Visualization of 4D Top Plans In the field, archaeologists should be able to visualize top plans as they are recorded with the ability to traverse in time how it looked in 3D during previous days. They should be able to visualize the data in its full resolution (no simplification or loss of fidelity) with viewable frame rates (min 24fps). This will require more efficient rendering systems that can run on a modern laptop in the field. Second, it requires more efficient algorithms for processing photogrammetry, syncing 3D scans, images, and other 3D recorded measurements. The software should be able to integrate seamlessly the data as it becomes available and dynamically update the top plans accordingly. Objective 4: Analytical Tools tailored for 4D Data Analyzing and manipulating data in a 4D environment will necessitate novel tools to facilitate annotation, editing, drawing, segmentation, and time-sensitive comparisons. In the remainder of this paper we present the current contributions to this new field with ArchField C++ software and how our field methodology has changed to account for this new approach. We present the results, challenges, and future directions. IV. E VOLUTIONARY ADVANCES: ARCHFIELD C++ The development of ArchField has been an evolutionary process since 2009, undergoing five excavation seasons of modification and adaption to meet new demands in multifaceted digital field excavations. Previous versions were designed to run as imbedded web applications interoperable with many GIS software tools and even deployable on handheld tablets and phones [22]. However, there were a number of drawbacks to this approach primarily related to the limitations of being locked inside a web browser, limited memory allocation, slow rendering speed, restrictions on access and control, and complicated installation of the web server system or iOS app. Until the