The point objectives in the hyper-spectrum image are difficult to be identified by the geometrical figure. It is needed to
reduce the dimension of the spectrum data in order to eliminate the Information superabundance of the multi-dimension
hyper-spectrum image data. We develop the distributed rules of the point objectives scatter plot in the low-dimension
space and confirm that the point objectives End Member are mainly distributed in the lesser confidence interval while
with the higher confidence coefficient. Finally we put forward to eliminate the non-point objectives and the noise End
Member going beyond the threshold so as to ensure the result of the characteristic clustering is effective. Based on the
scatter plot analysis, we find the new method to extract the spectrum characteristics by which we combine the
mathematics analytical models, statistical computing and the distinguishing effects tests. At the same time we establish
the model of spectrum character distinguish. According to the basic characteristics of the spectrum reflection features in
the green vegetation we confirm two kinds of characteristic bands, setting up the training type, and one-dimension vector
is formed after sampling by linearity combination. Through the practical application, we find the rather perfect spectral
classification characteristics and the discriminant function for both the point objectives and background.
During the course of image effect evaluation, random and fuzzy factors are both included. Traditional effect evaluation theory adopts classical probability statistical method, which only shows the random factors of remote sensing exploration system, and does not show the influential factors of psycho-physics and psycho-physiology, which are fuzzy factors, hard to be described and quantized exactly. The image effect evaluation method and computer simulation model are set up based on the interaction of mathematical model and probability statistics, and on the Hamming distance. The technology and thinking method can be applied to solve the problem of image effect evaluation in other similar systems.
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.