- 122 - 
vectors are plotted followed by further quality assurance (QA) 
procedures designed to ensure detail has been plotted faithfully 
and that the specified CAD layering system has been adhered to. 
The captured data can then be archived and digital drawings 
and paper plots prepared. The final stage is the delivery of the 
survey information in both paper and digital form (contained on 
a CD), probably via post. The originator of the work will then 
review the deliverables and perform a final QA phase. 
The introduction of new techniques for metric survey requires 
new workflows to be developed and the definition of new 
appropriate specifications. How a new technique compares 
with traditional techniques also needs to be considered. This 
requires the critical evaluation of both the methods and 
resulting data. For cultural-heritage recording any technique 
that is capable of producing data similar to the current products 
of floor plans, elevation plots and sections should be considered 
for its potential to improve the efficiency of survey data 
acquisition or improve the quality and value of the final product. 
Laser scanning is one technique that currently shows potential 
for being used for cultural-heritage recording. This paper 
describes the use of two different laser scanning systems on 
separate historic sites and describes the experiences of data 
acquisition and processing in comparison to traditional 
techniques. It also discusses how laser scanning would fit into 
a contemporary survey workflow. 
These projects were undertaken with a survey based approach 
to laser scanning, typified by the use of control data and the 
type of structures (e.g. subjects typical of the type that would be 
surveyed using current methods). The methodology used for 
the scanning in both projects can be likened to the 
photogrammetric flow-line described above. In addition to the 
collection of control data using terrestrial survey methods, the 
flow-line includes a registration phase (akin to stereomodel 
orientation) where the raw scans are brought onto a common 
coordinate system, data archiving phases and QA phases where 
the scan data is assessed for its quality. Unlike 
photogrammetry, however, no defined product exists for laser 
scanning and so for the projects described below the end- 
product was specified as point clouds registered to a common 
control scheme. The production of final end-products was then 
pursued following the data capture stage. 
2. CASE STUDY 1: ASHBY-DE-LA-ZOUCH CASTLE 
2.1. The Ashby Castle site 
Ashby Castle was originally converted from a wooden Norman 
fort to a more substantial stone building in around 1160. After 
several changes of ownership its possession reverted to the 
crown in 1461, followed in a few years by the construction of 
many domestic buildings and the largest structure on the site, 
Hastings Tower. During the English civil war the castle was 
captured by parliamentarian forces and, in common with other 
captured fortifications of the time, Hastings Tower was 
rendered unusable (Batt, 2000). Preservation work began in the 
19 th Century and the castle is now in the care of English 
Heritage. The south façade of the remains of Hastings Tower 
was selected as a test site to be laser scanned; it is 
approximately 24 metres in height and over 10 metres wide. A 
considerable amount of depth is present on this façade as it 
includes not only the interior wall but also the remains of two 
wall end-sections which protrude up to four meters from the 
main wall, as shown in Figure 1. The most appropriate method 
of recording this structure for redevelopment or conservation 
activities would be digital photogrammetry based on a local site 
coordinate system, producing rectified photography, 
orthorectified photography, or line drawings. In this project 
features of interest include a fireplace at the top of the tower 
which would be inaccessible by current recording techniques 
without specialist equipment - measurement of these areas by 
photogrammetry would require access via a hydraulic lift or 
scaffolding. The tower is considered by English Heritage a 
typical United Kingdom cultural-heritage site that would 
require close range recording. 
2.2. Laser scanning 
Figure 1. Hastings Tower, Ashby Castle. 
The scanner used at Ashby Castle was a Riegl LMS Z210 
instrument. It is a pulsed time-of-flight scanner with a 
maximum range of 450 metres depending on the reflectivity of 
the target (Riegl, 2002). In this project, emphasis was placed 
on a quantitative assessment of the instrument; therefore, 
scanning was performed at three different ranges from the face 
of the main façade 30m, 50m and 80m, rather than 
concentrating on collecting a complete description of the tower. 
As part of this evaluation process a large number of reflective 
targets were used to assess the precision and accuracy of the 
scanner - further details can be found in Mills and Barber 
( 2002 ). 
The remaining description considers the scans performed at 
30m (in this case providing a scan with an average resolution, 
on the main wall, of 50mm). The instrument is quoted as 
providing a single point accuracy of approximately +/- 26mm. 
The size of the laser footprint at 30 metres is 90mm. In 
addition to the capture of XYZ data the scanner also records a 
RGB value for each point using a one pixel CCD sensor which 
allows the scan data to be viewed as a colour image in addition 
to an intensity image, based on the strength of the returning 
pulse (Figure 2), or as a range image. Scanner control is 
provided by a laptop computer running proprietary software 
with the progress of the scan displayed as a 2D image while the 
scan is running. The Riegl system offers an averaging routine 
whereby several individual scans can be performed and used to 
create one averaged scan. In this case 12 scans were used to 
create an averaged result. The resulting scan had a field of 
view of 80 by 50 degrees and consisted of over 300 000 points. 
Registration of the scan to a common coordinate system was 
made possible through the use of retro-reflective targets which 
were located on the local site grid using theodolite intersection 
from two stations.