Remote sensing sensors are now able to deliver greatly increased amount of information with the used of high resolution
sensor. But high or very high resolution sensors lead to noise in generally homogeneous classes as the data contains
increased information with more internal variability. Conventional classification methods commonly cannot handle the
complex landscape environment in the image. The result of each method has often "a salt and pepper appearances"
which is a main characteristic of misclassification. It seems clear that information from neighboring pixels should
increase the discrimination capabilities of the pixel based measured, and thus, improve the classification accuracy and
the interpretation efficiency. This information is referred to as the spatial contextual information. In this paper, we shall
present a contextual classification method based on a frequency-based approach for the purpose of land cover mapping.
Additionally, classification maps are produced which have significantly less speckle error. In order to evaluate the
performances of the classifier, 9 different window sizes ranging from 3x3 to 19x19 with an increment of 2 is tested.
Traditional aerial images provided by satellite, manned aircraft or stock photography are often expensive, difficult to
obtain or outdated. The CropCam provides GPS based digital images on demand and real time data with high temporal
resolution throughout the equatorial region where the sky is often covered by clouds. The images obtained by the
CropCam will allow producers to detect, locate, and have better assessment of the actions required to overcome the
problem of unclear images obtained by the satellite and manned aircraft in this area. A Pentax digital camera, model
Optio A40, was used to capture images from the height of 320 meters on board the CropCam UAV autopilot. The
objective of this study is to evaluate the land use /land cover (LULC) features over Penang Island using the images
obtained during the CropCam flying mission. The study also test the effectiveness of neural network approach instead
of conventional methods in classification process in order to overcome or minimize the difficulty in classification of the
mixed pixel areas using high resolution images with spatial ground 8 cm. The technique was applied to the digital
camera spectral bands (red, green and blue) to extract thematic information from the acquired scene by using PCI
Geomatica 10.3 image processing software. Training sites were selected within each scene and four LULC classes were
assigned to each classifier. The accuracy assessment of each classification map produced was validated using the
reference data sets consisting of a large number of samples collected per category. The results showed that the neural
network classifier produced superior results and achieved a high degree of accuracy. The study revealed that the neural
network approach is effective and could be used for LULC classification using high resolution images of a small area of
coverage acquired by the CropCam UAV.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.