Dept, of Remote Sensing and Geoinformation, Military University of Technology, Kaliskiego 2, Str. 00-908 Warsaw,
Poland - (wdebski, pwalczykowski)@wat.edu.pl
KEY WORDS: Hyper spectral, Acquisition, Camera, CCD, Image, Video
ABSTRACT:
Remote sensing uses many different types of imagery for interpretational purposes. Multispectral photography has shown
to be of especially high interpretational value. During remote sensing studies which are done “in situ”, determining spectral
reflectance coefficients of any given object is crucial. Spectrometers are usually used in acquiring such data. The authors of such a
study have used an equipment set consisting of a digital video camera and an electronically tunable filter. The set is designed to work
in the visible range 400-720nm. The matrix in this camera has a resolution of 1392 x 1040Px. It is capable of binning (the joining of
pixels) 2x2, 4x4 and 8x8, which allows for the increase in sensitivity at the loss of resolution. Monochrofnatic images acquired by the
camera have a radiometric resolution of 12bits. The images are saved into the controlling computers memory by means of the IEEE-
1394 FireWire interface. The camera has a fully automated exposure mechanism together with an ROI function. The electronically
tunable filter allows for the acquisition of images of a lOnm band, with a O.lnm step, within the given spectral range. Special
software had been designed to control the set (the camera and filter), which allows for the acquisition of any chosen sequence of
images within the entire visible range of the electromagnetic spectrum, as well as their primary processing. On the basis of a
sequence of images, the spectral reflection coefficient can be determined for any objects within the scene. The image sequences can
be acquired in laboratory conditions, as well as during fieldwork.
The article describes the methodology in acquiring imagery using the video camera and the electronically tunable filter, as
well as the determining of the spectral reflection coefficients. Additionally a comparison will be made between the reflection
coefficients obtained by means of a spectroradiometer and those derived from the registered imagery.
1. INTRODUCE
In field measurements of light reflectance play an
increasingly important role in researching the environment using
remote sensing techniques. These are essential in both
determining an appropriate spectral band, which is to be
registered, as well as in aiding in the correct interpretation of
acquired multispectral imagery. Obtaining the maximum amount
of imagery about the value, character and variations of the
spectral reflection coefficients as well as their dependency on
lighting conditions guarantees a more complete image
interpretation. The measurements of the incident light and the
radiation reflected by the object at the moment of registration are
key. At the stage of planning a way for the acquisition of images
for remote sensing purposes, it is essential to know the
relationship which describes the contrast between the object and
the background. The registering system should measure its
maximum value. To ensure full remote detection it is
recommended that an appropriate channel is found, in which
there is a considerable difference between the reflection
characteristics of the object and the background - in other words
there is a high contrast between and object and the background.
During “in situ” studies, different types of spectroradiometers
are used to determine the reflectance characteristics of objects.
This however requires the use of complicated procedures to
obtain accurate results. Together with the appearance of
electronically tunable filters on the market, the authors propose
the use of a video camera with such a filter to obtain such
information.
2. THE SET DESIGN
The hyperspectral set, designed and created at the
Military University of Technology in Warsaw, consists of the
following elements:
- An electronically tunable narrow banded optical filter
- A monochromatic digital camera
- Software used to operate the hyperspectral acquisition of
images
Tuneable
filter and
control
module
lens
Digital
camera
i
k i
k