An Adaptive Noise-Filtering Algorithm for AVIRIS Data With Implications for Classification Accuracy
read more
Citations
The use of remote sensing in soil and terrain mapping — A review
Recent Advances on Spectral–Spatial Hyperspectral Image Classification: An Overview and New Guidelines
Communication systems: An introduction to signals and noise in electrical communication
Hyperspectral Image Classification Based on Structured Sparse Logistic Regression and Three-Dimensional Wavelet Texture Features
Dimension Reduction Using Spatial and Spectral Regularized Local Discriminant Embedding for Hyperspectral Image Classification
References
A transformation for ordering multispectral data in terms of image quality with implications for noise removal
Remote sensing, models, and methods for image processing
Numerical methods and software
Enhancement of high spectral resolution remote-sensing data by a noise-adjusted principal components transform
Communication systems: an introduction to signals and noise in electrical communication
Related Papers (5)
Classification of hyperspectral remote sensing images with support vector machines
Frequently Asked Questions (10)
Q2. What is the important reason for the large dimensionality of a hyperspectral dataset?
The large dimensionality of a hyperspectral dataset often requires a data transformation such as principal components analysis (PCA) or the singular value decomposition (SVD) to reduce the number of variables, or bands, within an image prior to further processing.
Q3. What is the advantage of the technique used?
An advantage of the technique used (discriminant analysis) is that individual bands are selected by the method, showing that high order noisy MNF bands contain signal that impacts applications such as classification.
Q4. What is the simplest way to estimate the noise in MNF coordinates?
In order to study the effects of filtration on SNRs in MNF coordinates, a estimate suggested by Schowengerdt is used:SNR = σ2S σ2N ,where σ2S is the variance of the signal and σ 2 N is the variance of the noise [15].
Q5. What is the definition of adaptive filter?
An adaptive filter can alter the size of the filter kernel (spatial domain) or change the frequencies filtered (frequency domain) depending on image characteristics and noise levels.
Q6. What is the eigenvalue of a b-band image?
The eigenvalues contained in Λ are the estimated variance of the signal (σS) divided by the estimated variance of the noise (σN ), and therefore the diagonal element λb in Λ is an approximation for the SNR of band b in the transformed image.
Q7. When was the AVIRIS 224-band imagery acquired?
Three flight-lines of Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) 224-band imagery were acquired in the winter of 1999.
Q8. How do you reduce the variance of a small window?
Spatial median filters work by decreasing the variance within a small window (kernel) by assigning a pixel the median value of the surrounding pixels.
Q9. What are the main reasons why hyperspectral images are noisy?
these images tend to be noisy as a result of the fine discretization and other factors such as the method of acquisition (small aircraft) [1].
Q10. How many bins can be divided into the MNF?
In order to divide the bands into bins in this manner, the area underneath the eigenvalue curve can be divided evenly into a number of bins corresponding to the number of different sized filters to be applied, as shown in Fig.