|
This extension is a project of Google Summer of Code 2010 for Opticks. The extension contains target detection, anomaly detection and data merge tools. |
Download 1.0 - September 26th, 2010 |
Summary
| License: | LGPL |
| Price: | free |
| Author: | Dalong Cheng |
| Website: | |
| Latest Version: | 1.0 updated on 2010-08-16 |
| Opticks Supported Versions: |
4.9.X |
| Operating Systems: |
Windows 32-bit |
| Support: | Send E-mail to yuguess@gmail.com if you need assistance |
Data Merge
This feature will help user to merge hyperspectral data from separate files. In this plugin, you could merge several one-band spectral data into a single multi-band spectral data. In order to merge the data, you firstly need to import the data through Opticks, the plugin will automatically list the imported data and let user to choose which of them should be merged. Be aware that all the merged data should have same row and column number. The user could also adjust the band order of
merged data through "up" and "down" button in the GUI.
Target Detection
The target detection will firstly define a concept to measure the similarities between target spectrum and background spectrum, then it will calculate each pixel. In this extention, I implement Adaptive Cosine Estimator(ACE) algorithm for target detection. Fisrtly, you should pick up a pixel in the Spectral image as target spectrum,
then the ACE will take this pixel as target and compare with all the pixel in the Spectral image. The result will be represented in a binary image.
Anomaly Detection
The anomaly detection algorithn will calculate band data of each pixel with its round pixels. If the difference above certain threshold, the pixel will be treated as a nomaly pixel, since band value it is not comform with its background. In this extention, I use Reed-Xiaoli algorithm for anomaly detection. Due to the large computation of the algorithm, the RX can only be applied to small data spectral data set. Just like ACE algorithm, the RX result will be displayed as a binary image.
References
Please refer to [1] and [2] for detail mathematics principles of RX and ACE algorithm.
[1] I.S. Reed and X. Yu, “Adaptive multiple-band CFAR detection of an optical pattern with unknown spectral distribution,” IEEE Trans. Acoustics, Speech, Signal Processing, vol. 38, pp. 1760-1770, Oct. 1990.
[2] S. Kraut and L.L. Scharf, “The CFAR Adaptive Subspace Detector Is a Scale-Invariant GLRT,” IEEE Trans. Signal Process. 47 (9), 1999, pp. 2538--2541
Screenshots and Videos
Figure 1 ACE test Image, I mark a red circle on the target I choose
Figure 2 The binary image of ACE result
Figure 3 RX test image
Figure 4 the binary image of RX result
Figure 5 The GUI of Data Merge Project
FAQ
All Download
All source code are available at http://github.com/yuguess/GSoC/