In: Wagner W„ Székely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
In: Wagi 
74 
THE MULTI-SENSOR LAND CLASSIFICATION SYSTEM 
LCS: AUTOMATIC MULTITEMPORAL LAND USE 
CLASSIFICATION SYSTEM FOR MULTI-RESOLUTION DATA 
A. Beccati* 3,b , M. Folegani 3 , S. D'Elia c , R. Fabrizi 3 , S. Natali 3 , L. Vittuari d 
3 MEEO srl, Research and Development, Via Saragat, 9,1-44122, Ferrara, Italy - (beccati, folegani, natali)@meeo.it 
b University Of Ferrara, Department of Mathematics, Via Saragat, 1,1-44122, Ferrara, Italy - alan.beccati@unife.it 
c ESA - ESRIN, Ground Segment Department, via Galilei 1,1-00044, Frascati, Rome, Italy 
d University of Bologna, DISTART, Viale Risorgimento 2,1-40136, Bologna, Italy 
Technical Commission VII Symposium 2010 
KEY WORDS: Land Cover, Land Use, Modelling, Web based, Global, Multiresolution, Multi temporal, System 
ABSTRACT: 
Providing land use/land cover change maps through the use of satellite imagery is very challenging and demanding in terms of 
human interaction, mainly because of limited process automation. One main cause is that most of land use/land cover change 
applications require multi-temporal acquisitions over the same area, that introduces the need for accurate pre-processing of the 
dataset, in both geo-referencing and radiometry. Moreover, single multi-spectral images can be hundred of megabytes in size and 
therefore image time series are even more difficult to be handled and processed in real time. The approach here proposed foresees 
the use of a robust land cover classification system named SOIL MAPPER® to reduce input data size by assigning a semantic 
meaning (in the land cover domain) to each pixel of a single image. Land cover transitions and land use maps can be expressed as 
evolutions of land cover classes (features) on temporal domain. This permits to define ‘trajectories’ in the features - time space, that 
define specific transition or periodic behaviour. The target system, named Land Classification System, provides fully automatic and 
real time land use/land cover change analysis and includes fundamental sub-systems for accurate radiometric calibration, accurate 
geo-referencing (with geolocation within the pixel size) and accurate remapping onto an Earth fixed grid. The characteristics of the 
selected pre-classification system and Earth fixed grid allow general application across different sensors. Land Classification System 
has been prototyped over 15 years of global (A)ATSR data and foresees integration of over 3 years of regional ALOS-AVNIR-2 
data. 
1. INTRODUCTION 
Land use and land cover change (LULCC) topics are going to 
become more and more critical subjects for the impact they 
have on the global climate. They are in fact linked to climate 
and weather in complex ways and are fundamental inputs for 
modelling greenhouse gas emissions, carbon balance, natural 
ecosystems and human environment evolution. Both human 
activity and natural phenomena can affect many of these 
processes, that are strictly correlated, influence each other and 
have strong impact and consequences on environmental, social 
and economic aspects as well as on human health. Land cover 
refers to everything that covers the land surface, including 
vegetation, bare soil, buildings and infrastructure, inland bodies 
of water, and wetlands. Land use refers, instead, to societal 
arrangements and activities that affect land cover (Mahoney et 
al„ 2003). 
Local-to-global scale LULCC studies and application has got 
great benefits from the use of remotely-sensed data, mainly due 
to the preferred point of view of satellite platforms for the 
periodic monitoring of the territory. Besides existing long time 
series of satellite data archives, there is an ever increasing 
availability of satellite images with global coverage from 
different sources and at different resolutions. As a drawback, 
single multi-spectral image can be hundred of megabytes in size 
and the real time utilization of these datasets for online analysis 
is a technological challenge by itself; that, paired with the high 
amount of time required for semi-automated image analysis 
suggests that fully automated pre-processing systems shall be 
used to improve satellite data exploitation and reduce the data 
volume at the same time. 
In order to improve multitemporal satellite data usability for 
time series analysis, accurate image pre-processing operations 
shall be performed to make time series datasets homogeneous: 
the most important pre-processing steps are accurate 
geolocation and accurate radiometric calibration; digital 
numbers to radiance or surface reflectance conversion must be 
performed for quantitative analyses of multi-temporal images 
(Lu et al., 2004). Precise geolocation and image registration are 
to be addressed on a per-sensor basis, since each one has its 
geometric properties and correction factors. 
The proposed generalised approach, named Land Classification 
System (LCS), aims at exploiting advanced applications for 
single image feature extraction, providing easy-to-use tools for 
land use and land cover change detection analysis over time 
series of data; such approach, to be readily usable by the 
scientific community and also by end users of land cover and 
land use maps, is also aimed at providing a computer aided 
modelling and land cover evolution analysis tool for definition 
of evolution models by domain experts and an high degree of 
automation for evolution models application in an effort to ease 
and speed-up the analysis of land use and land cover change 
phenomena, possibly in conjunction with other tools to find 
correlations among different factors influencing life on Earth 
like global climate. 
In the fc 
impleme 
the critic 
LCS bei 
Space A 
Applicat 
obtained 
aim of d( 
1.1 Rel 
Many a] 
change a 
2004). S 
been perl 
develop 
interestir 
is provid 
methodo 
comprehi 
available 
data, anc 
between 
(Haines-' 
is not foe 
or pheno 
land covi 
maps sto 
for scien 
their rese 
makers b 
provide f 
The LCS 
satellite ( 
The tern 
system 
permittin 
frames; i 
multitem] 
analysis. 
LCS crea 
use of a 
MAPPEF 
different 
classifica 
permittin: 
The stock 
reference 
analysis 
amounts 
semantic 
Terabytes 
modern o 
The core 
evolution 
evolution 
classes a 
temporal