3D Modelling Of A Famosa Fortress, Malaysia
Based on Comparison of Textual and Visual Data
M.Izani, A. Bridges
Dept. of Architecture
Strathclyde University
131 Rottenrow G4 0NG UK
(mohamad.zainal-abidin , a.h.bridges)@strath.ac.uk
Abstract— This paper presents an attempt to model the “A
Famosa Fortress” in Malaysia into 3D. This building was built
in 1511 by the Portuguese and went through several
architectural developments and changes before being largely
destroyed during the British occupation in 1824. The biggest
challenge in this research is to determine the original fortress
layout due to the lack of any authoritative documentation
pertaining to the fortress. Detail analysis has been conducted
to identify reliable sources for references which are available
in the form of text and visual. In this paper, we focus on
comparison of selected textual and visual data to come out with
a verifiable conjectural layout of the fortress. We then pre-
visualized the layout in 3D model. Some samples of the model
are presented here however there are still rooms for
improvements before it is finalized. The output of this research
will be tested for application in tourism and education.
Keywords- A Famosa,3D model, fortress
I. INTRODUCTION
Historically, Melaka’s strategic position in South East
Asia has made it as an important centre of commerce [1].
The traders used to trade various items such as spices, cloth,
tin, silk, porcelain and many more. These traders came from
all over the world: India, China, Borneo, Arabia and Europe.
Melaka’s popularity attracted the Portuguese to expand their
power in commercial dealings, military occupation and
religion. The Portuguese believed that by controlling
Melaka, they could monopolise spice trading which was a
very valuable item in Europe and expand their military
power. One said that whoever is lord of Melaka has his hand
on the throat of Venice [2]. Besides, one of the Portuguese’s
objectives was to expand the influence of Christianity in this
region and this could only be done by seizing Melaka. In
1511, the Portuguese, with fifteen small and great sails and
with sixteen hundred fighting men laid siege to Melaka [3].
With advanced strategy and weapons, the Portuguese
managed to capture Melaka within three weeks and, on
August 10th, 1511 Melaka fell into Portuguese hands [4].
Albuquerque was the captain for the new Portuguese
government in Melaka. He immediately ordered a fortress to
be built for defensive purpose [5]. With this success, it also
attracted the Dutch with the monopoly of the trading in
South East Asia. The fortress of Melaka continued its
A. Razak
Faculty of Information Technology
Multimedia University
Cyberjaya, Selangor, Malaysia
aishah@mmu.edu.my
architectural development over this time. Prior to the
occupation of the Dutch, they heavily bombarded the fortress
which has critically destroyed part of the fortress. After they
succeeded to conquer Melaka in 1641, the Dutch carried out
major reconstruction on the fortress as part of their strategy
to strengthen their power. This reconstruction involved the
extension of the fortress walls and bastions [6]. When the
British took over the Dutch’s position in 1824, the British
captain in Melaka, William Farquhar instructed the fortress
to be destroyed. As the result, the only evidence left today is
a gate to access the fortress which is known as Porta de
Santiago (Figure 1).
Figure 1. Remnant of Old Portuguese fortress in Melaka, Porta de
Santiago [7]
A. Motivation
In 2003, the foundation of the Santiago bastion, part of
the walls of the Fortress Melaka was discovered during
piling works for the Dataran Pahlawan Melaka Megamall
project in Bandar Hilir. In November 2006, while the
Melaka Government was building the 120 meters high
“Melaka Tower” project just inside the heritage zone,
besides the Melaka river, the piles of the tower hit something
hard below the ground. Only upon excavation, the workers
discovered the walls and foundations of another fortress
bastion identified as Middleburgh. This new discovery was
so important that the Ministry of Culture, Arts and Heritage
has allocated more than RM 12 million to excavate and do
research of this new founding. However, work to uncover
more of the buried wall, despite its historical significance
and potential to draw more tourists into the area, may not be
feasible as it would involve tearing up the major road that Figure 2 gives detail plan of the Santiago date extracted from
links the old town with the new commercial area. Hence, the original site.
only certain part of the fortress can be reconstructed. In order
to solve this problem, we propose to digitally reconstruct this
fortress in form of 3D model which can be navigated in the
virtual world. Since the fortress itself has faced several
changes in its design and layout, by reconstructing it in 3D
allows the researchers to investigate and study the
development of these changes from architectural and
historical aspects.

II. DATA COLLECTION

After getting our resources, we started to extract the
relevant data and classified it into different forms. They are
listed below.
A. Visual data such as drawings, paintings and map
We have extracted and collected a lot of visual data
related to A Famosa from books, journals and even direct
from the national archives. Among them are old drawings
dated from year 1500 to 1700 recorded by various artists,
draftsmen and architects from the Portuguese era until the Figure 2. Detail plan of Porta de Santiago [9]
British era. Unlike the drawings from the Portuguese era
most of the drawings during Dutch era are more precise. However, after years of excavation works, a bastion
They are in the form of plans, drawn according to scale with known as Middleburgh has been successfully reconstructed.
specific measurement unit. However, the measurement unit Figure 3 gives the detail plan of the Middleburgh bastion,
used are old units such chain, rod, furlong and fathoms received from the excavation team.
which are no longer used nowadays. Besides that, even the
same unit might use different conversion value such as listed
in Table 1 below.
TABLE I. CONVERSION OF OLD UNITS [8]

Name Conversion (1 unit equivalent)

11 Dutch inches
Dutch foot
11 3/8 English inches
English foot 12 inches
Dutch rod Sometimes 11,12 13 Dutch feet
English rod 16.6 English feet
Figure 3. Detail plan of the Middleburgh bastion, received from the
excavation team.
B. Textual descriptions
At the moment, the excavation team is working on
Drawings extracted from the reports, books, journals, reconstructing another bastion, known as Frederick
paper and etc are usually further explained in textual forms. Hendrick. The data we obtained from the physical inspection
Even though the information is quite scattered compared to of the fortress can be used as our main reference in our
the visual from, proper extraction can give important analysis and can be projected to the findings from visual and
information on the fortress material, height, position, textual data in order to get a complete accurate data of the
elements and functions which can give better understanding whole fortress wall.
of the fortress. This can be an important reference and
support data for translating the visual data.
III. DETERMINATION OF FORTRESS LAYOUT
C. Physical data
The physical data of the fortress is the most accurate data A. Translating textual description into visual
to represent the fortress. Initially, the only available remain Based on the collected information, we tried to construct
of the fortress after the British era is Porta de Santiago, a verifiable conjectural layout of A Famosa fortress. In our
which was one of the gates to the fortified city of Melaka. first approach, we have tried to construct the layout based on
the report of Governor Balthasar Bort in Malacca dated 1678
[10] and description given by Manuel Godinho De Eredia in
1604 [5]. A detail analysis has been done to compare and
verify the given data [11]. Based on the findings, we have
constructed the A Famosa layout as in Figure 4.
Figure 4. Complete visualization of innermost polygon and the outermost
line of A Famosa including the bastions
B. Procrustes analysis
To compare the layout that we have constructed in Figure
4 with the visual data that we have collected, we decided to
apply Procrustes analysis approach. Procrustes analysis is a
rigid shape analysis that uses isomorphic scaling, translation,
and rotation to find the “best” fit between two or more
landmarked shapes. In our case, we perform a least-squares
orthogonal generalized Procrustes analysis (least-squares
orthogonal mapping). It applies nonlinear mapping algorithm
to find best fit, i.e. to find a reference cluster of landmarks,
so that the distance of each reference landmark to it's
corresponding experimental landmark is minimized. Detail
discussion on this approach is presented in [12]. Based on
the result from Procrustes analysis, we decided to use Map A
and Map B (out of 12 maps selected earlier) as the main
visual reference in our attempt to model this fortress in 3D.
These maps are presented in Figure 5.
Map A Map B
Title: Map of the city and fort at Title: Plan of Malacca fort, ca.
Malakka, 1780 [13] 1792, (after Reimer and Elias)
[14]
Figure 5. Selected map for visual reference
Information that we get from the map is very helpful in
modeling the 3D model of A Famosa. This is because it
visualizes the bastions shape and size, the fortress layout and
other details, which is not available in the textual report or
description. However, the visual data, might not give the
most accurate information in terms of measurement [15, 16].
Thus in order to get the most accurate measurement for our
model, comparison of visual and textual is necessary.
C. Comparison of measurement from textual and visual
data
In the next step, we intended to compare point to point
measurement of the visual data (Figure 4, Map A and Map
B) to the textual data (Balthasar's report and Eredia’s
description). One problem for this task is to determine the
start and end point for the measurement. The reason for this
is because we do not know where is the start and end point
for the measurement given by the textual data. Thus to
reduce the error resulted from different start and end points,
we have done some trial and error measurement with several
combination of start and end point. Based on the findings,
we come out with the points given Figure 6.
Figure 6. Rough point to point measurement
Another attempt to reduce this problem is to reduce the
number of measurement points. This is done by grouping the
fortress into 2 groups. Group 1 consists of the part of the
fortress which has been modified during the Dutch
occupation. This group consists of the measurement from
Frederick Hendrick to the bastion of Emelia. The
circumference for this group is denoted as L1. Group 2
consists of the part of the fortress which has not been
modified during the Dutch occupation and it consists of the
measurement from Emelia to the bastion of Frederick
Hendrick. The circumference for this group is denoted as L2.
Two readings are made from each measurement which are:
1) Detail point to point measurement connecting each
landmarks in Figure 6 (considering every curve and corner).
2) Rough point to point measurement, connecting each point [10], he reported that some modifications were done on the
as shown in the outermost line in Figure 6. The reason for fortress during his time such as the bastion of Victoria was
taking the rough measurement is based on our assumption extended to 1/3 of it’s original size and the addition of
that measuring device at that time is limited and detail Middleburgh bastion, which is not exist during the
measurement might not be available. This assumption is Portuguese era. Further analysis shown that the measurement
further supported by the fact that both Balthasar's report and reported in Balthasar's report is smaller compared to the
Eredia’s description does not describe in detail the size of value calculated from the visual description. One explanation
each bastion. for this is given by [14]. He stated that measured drawings
may be rare for sixteenth century machine drawings,
fortifications plans are littered with the numbers and
IV. RESULTS AND ANALYSIS calculations, from the preparatory to the final drawings.
Figure 7 presents the comparison of point to point Early engineering drawings from the 14th and 15th centuries
measurement of textual and visual data. Even though we can have been preserved almost exclusively in the context of the
see that the value shows some variation for different sources, production of presentational manuscripts. All of the
in general it can be seen that the variation for Group 2 is drawings of the 15th and 16th centuries may appear
small compared to Group 1. This confirms that the data for unprofessional to a modern beholder. They in no way
this part of the fortress is consistence in the visual and resemble the orthographical plans and schematics that
textual form thus there is no need for further clarification. engineers trace and employ today. This also explains why
the rough measurement of the visual data is closer to the
value reported in Balthasar's report compared to the detail
1800 measurement.
Balthasar's report
Further explanation is found in [13] which stated that
Eredia's description most of the early drawings, whether project drawings or
1600
Figure 4 detail memory records, could not have been working plans. The
drawings had limited utility as practical guides, lacking
Figure 4 rough
1400 technical specifications such as scale and jointing stones. At
Map A detail the beginning of work on most medieval buildings the
Map A rough architect was probably in a position to give his patron only a
general notion of what his building look like on completion.
1200 Map B detail He worked out details subsequently, in the course of
Map B rough construction; meanwhile he had to be in steady attendance to
1800
Balthasar's report
instruct the workers so they could realize his intentions,
1000
1600 Eredia's description which he carried around only in his head. Based on this
1400 Figure 4 detail explanation, we decided that the textual descriptions might
1200 Figure 4 rough be more accurate in terms of measurement compared to the
800
1000
Map Adetail visual descriptions, thus it will be our main reference. This is
Map Arough
because as explained previously, drawing technology during
800
that time is limited and merely dependent on the skills of the
Map B detail
600600 artists. However the measurements calculated from the
Map B rough visual descriptions (map A and map B) can be a guide in
400

verifying the conjectural layout that we have produced in

Mauritius to Frederick Hendrick
Henriette Louijse to Wilhelmus
Fredrick Hendrick to Victoria

200
Emelia to Henriette Louijse

Figure 4.
Wilhelmus to Mauritius

400
Total circumference
Victoria to Emelia

0
L2
L1

200
V. CONVERTING THE LAYOUT TO 3D MODEL
Based on the verified conjectural layout of A Famosa that
we have constructed in Figure 4, and the visual data from
0 Map A and Map B, we have started to model the layout into
3D. Figure 8 and Figure 9 present the top view and
perspective view of the whole A Famosa in 3D respectively.
In these two figures, we have included the element of sea
and land surrounding the fortress and also the hilly terrain
inside of the fortress wall.
Figure 7. Comparison of point to point measurement of different sources
Figure 10 presents the complete 3D modeling of Porta de
For the case of Group 1, we can see that Eredia's Santiago, shown in Figure 1. This is the only remnant of the
description gives the smallest value compared to other fortress which is still available today, thus the easiest part to
sources, including the Balthasar's report. This is expected model. Figure 11 illustrates the bastion of Middelburgh
due to the modification that was made during Dutch era as which is currently being reconstructed on site by the
mentioned in the Balthasar's report. In Balthasar's report conservation team. To complete the detail of the other part of
the fortress need further study which include visits to other
Portuguese fortress which are still available today such as
those in Goa, India and Sri Langka, Bangladesh. In this
research, we intend to focus on the fortress wall only, Detail
study on the material, elements on the wall such as gates,
irons, breastwork and etc are necessary in order to complete
the 3D modeling of the fortress.

Figure 10. 3D modeling of Porta de Santiago of A Famosa fortress in

Melaka, Malaysia

Figure 8. Top view of A Famosa Fortress in 3D

Figure 11. 3D modeling of Middelburgh bastion

VI. CONCLUSION AND FUTURE WORKS

Figure 9. Perspective view of A Famosa Fortress in 3D
Since the remains of the fortress are very minimal, the
only appropriate reconstruction method for this research is
traditional 3D polygonal modeling. The use of advanced
methods such as photogrammetry and 3D laser scanning are
not practical due to these methods requiring physical objects
to be measured. Thus, the biggest challenge in this research
is to come out with a verifiable conjectural layout of A
Famosa. This is due to the lack of any authoritative
documentation pertaining to the fortress. Extensive
verification approaches have been adopted to verify all
collected data and to determine the most accurate layout for
our reference in modeling the 3D model of A Famosa. Some
of the developed models have been presented in this paper.
However, there are still many details to be added to the
model such as the elements of the fortress walls and the
material of the wall. Once we have completed the model, we
would run some human evaluation test to identify the
accuracy and efficiency of the model to be used in education
and tourism industry.
REFERENCES
[1] Thomaz, Luís Filipe F. R., and Manuel Pintado. 2000. Early
Portuguese Malacca. [Macau]: CTMCDP/IPM, pp. 19.
[2] Pires, T., & Rodrigues, F. (1944). The Suma oriental of Tomé Pires,
an account of the East, from the Red Sea to Japan, written in Malacca
and India in 1512-1515, and The book of Francisco Rodrigues, rutter
of a voyage in the Red Sea, nautical rules, almanack and maps,
written and drawn in the East before 1515. London: The Hakluyt
Society.
[3] Ryan, N. J. (1960). Malaya through four centuries an anthology,
1500-1900. London, Oxf. Univ. Press, pp. 6
[4] Noonan, L. A. (1989). John of Empoli and his relations with Afonso
de Albuquerque. Lisboa: Ministério da Educacã̧ o, Instituto de
Investigacã̧ o Científica Tropical.
[5] Godinho de Eredia, M., & Mills, J. V. (1997). Eredia's description of
Malaca, Meridional India, and Cathay. Kuala Lumpur, Malaysia:
Malaysian Branch of the Royal Asiatic Society.
[6] Leupe, P. A., & Hacobian, M. (1936). The siege and capture of
Malacca from the Portuguese in 1640-1641: extracts from the
archives of the Dutch East Indies Company. Singapore, Malayan
[7] Asia Explorers., 2003. Article on Exploring the Porta de Santiago (A
Famosa) Remains of old Portuguese fort, Malacca, Malaysia.
http://www.asiaexplorers.com/malaysia/ porta_de_ santiago.htm
(accessed 02 Jun. 2008)
[8] Murray Hoffman (1862). Treatise Upon the Estate and Rights of the
Corporation of the City of New York, E. Jones & Co.
[9] Viana de Lima, A. E. (1988). Reviver Malaca = Malacca, a revival.
Porto, Viana de Lima e Livraria Figueirinhas.
[10] Bort, B. (1927). Report of Governor Balthasar Bort on Malacca,
1678. Singapore: Malayan Branch, Royal Asiatic Society.
[11] Izani, Bridges & Razak, A (2009), “Determination of the plan of the
A Famosa fortress, Malaysia”, to appear in proceeding of
International Conference on Computing and Informatics 2009, June
2009, in press.
[12] Izani, Bridges & Razak, A (2009), “Using procrustes analysis to
determine verifiable conjectural layout of A Famosa fortress,
Malaysia, to appear in proceeding of International Conference on
Software Technology and Engineering 2009”, July 2009, in press.
[13] National Archive, "Map of the city and fort at Malakka," [Online].
Available: http://www.nationaalarchief.nl/amh/ detail. aspx?page=
dafb&lang=en&id=2165#tab0. [Accessed: Sept. 12, 2008].
[14] Harrison, B. (1986). Holding the fort, Melaka under two flags, 1795-
1845.Kuala Lumpur, MBRAS
[15] Goldthwaite, R. A. (1980). The building of Renaissance Florence: An
economic and social history. Baltimore: Johns Hopkins University
Press.
[16] Lefe'vre, W. (2004). Picturing machines 1400-1700. Cambridge,
Mass, MIT Press.