IMPROVED REMOTE SENSING AND GIS RELIABILITY DIAGRAMS, IMAGE 
GENEALOGY DIAGRAMS, AND THEMATIC MAP LEGENDS TO ENHANCE 
COMMUNICATION 
John R. Jensen 
Sunil Narumalani 
Department of Geography 
University of South Carolina 
Columbia, SC 29208, U.S.A. 
ISPRS Commission VI 
ABSTRACT 
Students, scientists, and users viewing images or maps produced 
from remotely sensed imagery or GIS technology encounter a 
bewildering array of unstandardized, often uninterpretable map or 
image annotation. This research focuses on the development of 
improved image and thematic map annotation which is designed 
to enhance the reader's ability to extract information efficiently 
and accurately. The specific types of legend information under 
consideration include: 1) calibrated, gray-scale step-wedges 
surrounding images (and photomaps) to ensure correct exposure 
and visual presentation, 2) detailed image coordinate (image space) 
and map coordinate (map space) information including the map 
graticule, 3) improved thematic map legends for a) maps derived 
from individual dates of imagery, and b) change detection maps 
derived from multiple dates of imagery, 4) spatial reliability 
diagrams summarizing the thematic as well as geometric accuracy 
of sources of information used in the creation of the final image 
or thematic map, and 5) methods of storing and summarizing the 
lineage (genealogy) of each image or final thematic map. In 
addition, the process of developing such annotation should be 
facilitated by the incorporation of graphical user interfaces (GUIs) 
in all remote sensing and GIS software packages. It is believed 
that these improvements will increase the reader’s ability to 
understand what is portrayed in a map or image map. 
KEY WORDS: Remote sensing, User interface, Gray-scale step- 
wedges, Static/dynamic legends, Spatial reliability diagrams, 
Lineage, Genealogy. 
INTRODUCTION 
Remote sensing and geographic information systems (GIS) are 
rapidly growing technologies (Jordan, 1992). Applications of 
these technologies range from long term planning by federal, state 
and local governments to short term crisis management (e.g. 
Jensen et al., 1990; Graff, 1991; Albers et al., 1991; Narumalani, 
et al, 1992). The number of users of products from these 
technologies is also increasing. Unfortunately, the users often 
have little or no experience with remote sensing and GIS, and are 
only exposed to the final product, which in many cases is a 
thematic map or image map. Poorly designed maps restrict the 
communication of information, or may convey false impressions 
(Weibel and Buttenfield, 1988). Therefore, it is essential that the 
remote sensing and GIS community make a concerted effort to 
provide thematic maps and/or image maps that are a) 
geometrically and thematically accurate, and b) annotated using 
correct cartographic principles which are conducive to easy 
interpretation and comprehension. 
Lunetta et al. (1991) emphasize that decisions based on 
geometrically and thematically inaccurate maps and image maps 
increase the probability of implementing bad decisions. Many of 
the remote sensing and GIS products are often inaccurate because 
they do not meet basic cartographic standards. Cartographers 
emphasize the terms "clarity", "readability", and "aesthetics" in 
map design, construction, and reproduction (Crawford, 1971). 
125 
Unfortunately, remote sensing and GIS analysts often have little 
cartographic training. They must avail themselves of this body of 
knowledge to develop software and high quality output which 
incorporate correct cartographic principles. This paper focuses on 
the development of improved image and thematic map 
cartographic annotation which is designed to enhance the reader's 
ability to extract information efficiently and accurately. 
BASIC CONCEPTS 
There are six (6) basic considerations which, if properly 
implemented, can increase the probability of producing accurate 
and interpretable image maps or thematic maps. First, it is 
important that the image processing or GIS software being used 
has an effective user-interface which can facilitate the cartographic 
development of a final product. Second, there should be 
calibrated gray-scale step-wedges (or a color wheel or color bar) 
surrounding images (and photomaps) to ensure correct visual 
presentation on the CRT screen or *hard copy' output. Third, the 
presence of image coordinate (image space) and map coordinate 
(map space) information is essential to make the user aware of the 
coordinate system to which the map has been transformed. 
Fourth, it is important to provide accurate and improved thematic 
map legends for a) static maps, (those derived from individual 
dates of imagery), and b) dynamic or change detection maps 
derived from multiple dates of imagery. Fifth, the user should be 
able to refer to "spatial reliability diagrams" that summarize the 
thematic as well as geometric accuracy of the sources used in the 
creation of the final map product. Finally, since it is inevitable 
that several iterations have been performed on the data, there is an 
urgent need to have methods of storing and summarizing the 
‘lineage’ or ‘genealogy’ of each final image or thematic map 
product. It is instructive to review the nature and utility of each 
of these topics. 
USER INTERFACE 
The quality of the ‘user interface’ not only affects the ease-of-use 
of the software, but also contributes to the cartographic design of 
the final product (Driver and Liles, 1989). However, according to 
Barr (1986), and Cowen and Love (1988), effective user interfaces 
have not been strong points of GIS or remote sensing digital 
image processing systems, thus making such systems difficult to 
use for interactive map design. 
Basically, there are two types of user interfaces: 1) command line 
interfaces, and 2) graphical user interfaces (GUIs). Historically, 
user interfaces were simple command line prompts, where input 
(an executable command followed by parameters) was expected 
from the user (Mark and Gould, 1991). An improvement over this 
system was the development of "fill-in forms", which prompted 
the user to add required parameters for executing a function. 
These interfaces are still evident in some of the most widely used 
image processing and GIS software packages. The major 
drawback of the command line interface is the need for a user to 
memorize or continuously refer to a manual in order to correctly 
execute a function.