Towards the Concurrent Execution of Multiple Hyperspectral Imaging Applications by Means of Computationally Simple Operations
"> Figure 1
<p>FLow-chart of the proposed algorithm stages for each image block <math display="inline"><semantics> <msub> <mi mathvariant="bold">M</mi> <mi>k</mi> </msub> </semantics></math>. → * means that these variables are reused in other stages of the algorithm. If <math display="inline"><semantics> <mrow> <msub> <mi>p</mi> <mi>c</mi> </msub> <mo>></mo> <msub> <mi>p</mi> <mrow> <mi>A</mi> <mi>D</mi> </mrow> </msub> </mrow> </semantics></math>, then <math display="inline"><semantics> <msub> <mi>E</mi> <mrow> <mi>A</mi> <mi>D</mi> </mrow> </msub> </semantics></math> is a subset of <math display="inline"><semantics> <msub> <mi>E</mi> <mi>c</mi> </msub> </semantics></math> and vice-versa.</p> "> Figure 2
<p>Spectral response of the Specim FX10 hyperspectral camera. The range of wavelengths collected by the data set employed in this manuscript is enclosed between dashed lines.</p> "> Figure 3
<p>Google Maps pictures of the farming areas corresponding to the hyperspectral images that are used in this work. (<b>a</b>) Location of the terrains on the island of Gran Canaria. (<b>b</b>) Area covered during the first flight campaign over a banana plantation. (<b>c</b>,<b>d</b>) Area covered during the second and third flight campaigns over different vineyards.</p> "> Figure 4
<p>RGB representation of the hyperspectral data acquired in each mission campaign swath that was used in this work. Color squares highlight the regions selected for the experiments. (<b>a</b>) Image 1; (<b>b</b>) Image 2; (<b>c</b>) Image 3; (<b>d</b>) Image 4; (<b>e</b>) Image 5; (<b>f</b>) Image 6.</p> "> Figure 5
<p>RGB representation of the employed test bench. Pixels enclosed in blue circles represent the anomalous entities to be detected. (<b>a</b>) Image 1; (<b>b</b>) Image 2; (<b>c</b>) Image 3; (<b>d</b>) Image 4; (<b>e</b>) Image 5; (<b>f</b>) Image 6.</p> "> Figure 6
<p>Example of spectral signatures corresponding to some real pixels of the first image used in the experiments (<a href="#remotesensing-12-01343-f005" class="html-fig">Figure 5</a>a). (<b>a</b>) Pixel locations; (<b>b</b>) Pixel spectral signatures.</p> "> Figure 7
<p>Anomaly detection results for input parameters <math display="inline"><semantics> <msub> <mi>N</mi> <mrow> <mi>b</mi> <mi>i</mi> <mi>t</mi> <mi>s</mi> </mrow> </msub> </semantics></math> = 12, <math display="inline"><semantics> <mrow> <mi>B</mi> <mi>S</mi> </mrow> </semantics></math> = 1024 and <math display="inline"><semantics> <mrow> <mi>C</mi> <mi>R</mi> </mrow> </semantics></math> = [12,16,20]. Lines in blue color represent the <math display="inline"><semantics> <msub> <mi>n</mi> <mi>f</mi> </msub> </semantics></math> frames employed to estimate the background distribution. Green lines represent the hyperspectral frames free of anomalies. Red lines represent those frames identified by our proposal to be corrupted by anomalous pixels. Pixels enclosed in red circles represent the anomalous pixels detected by our proposal. (<b>a</b>) Image 1; (<b>b</b>) Image 2; (<b>c</b>) Image 3; (<b>d</b>) Image 4; (<b>e</b>) Image 5; (<b>f</b>) Image 6.</p> "> Figure 8
<p>Block Diagram of the processes involved in the proposed method (blue arrow) versus those involved when images are on-board compressed, and the anomaly detection process is off-line performed (red arrow).</p> "> Figure 9
<p>Anomaly detection results using compressed–decompressed images for input parameters <math display="inline"><semantics> <msub> <mi>N</mi> <mrow> <mi>b</mi> <mi>i</mi> <mi>t</mi> <mi>s</mi> </mrow> </msub> </semantics></math> = 12, <math display="inline"><semantics> <mrow> <mi>B</mi> <mi>S</mi> </mrow> </semantics></math> = 1024 and <math display="inline"><semantics> <mrow> <mi>C</mi> <mi>R</mi> </mrow> </semantics></math> = [12, 16, 20]. Lines in blue color represent the <math display="inline"><semantics> <msub> <mi>n</mi> <mi>f</mi> </msub> </semantics></math> frames employed to estimate the background distribution. Green lines represent the hyperspectral frames free of anomalies. Red lines represent those frames identified by our proposal to be corrupted by anomalous pixels. Pixels enclosed in red circles represent the anomalous pixels detected by our proposal. (<b>a</b>) Image 1; (<b>b</b>) Image 2; (<b>c</b>) Image 3; (<b>d</b>) Image 4; (<b>e</b>) Image 5; (<b>f</b>) Image 6.</p> "> Figure 10
<p>Reduction in the number of FLOPs (%) using the proposed methodology compared with the execution of both the lossy compression and the anomaly detection processes as two independent algorithms.</p> ">
Abstract
:1. Introduction
2. Materials and Methods
2.1. Proposed Set of Core Operations
2.1.1. General Notations
2.1.2. Gram-Schmidt Method
Algorithm 1 Modified version of the Gram–Schmidt Orthogonalization |
Inputs: |
Outputs: |
{Orthogonalized vector}; {Orthonormalized vector}; |
Algorithm: |
|
2.1.3. Set of Core Operations
Algorithm 2 Set of core operations |
Inputs: |
Outputs: |
{Average Pixel}; {Characteristic pixels}; {Orthogonalized vectors}; {Orthonormalized vectors}; {Projection vectors} |
Algorithm: |
|
2.1.4. Discussion about the Proposed Set of Core Operations
2.2. Proposed Methodology for Lossy Compression and Anomaly Detection
2.2.1. Set of Core Operations for Lossy Compression and Anomaly Detection
2.2.2. Subsequent Stages for Lossy Compression
2.2.3. Subsequent Stages for Anomaly Detection
Algorithm 3 Alternative method to compute the orthogonal projection matrix |
Inputs: |
Algorithm: |
|
2.2.4. Computational Complexity of the Proposed Methodology
- In this case, the set of core operations is used to extract the and most different pixels in each received image block, . For simplicity, we consider that is equal to in this analysis. In addition, the Preprocessing and Entropy Coding stages belonging to the compression process are also executed in order to codify the outputs of the core operations before being sent.
- On one hand, the set of core operations and the Preprocessing and Entropy Coding stages are applied to perform the compression of the image block at issue. On the other hand, the extraction of the purest background reference vectors within is also needed by the anomaly detection process as explained in Section 2.2.3. Concretely, this stage is referred as to Extraction of the Background Reference Vectors in Figure 1. For doing so, the set of core operations is used once again but in this case, plays as input matrix . Whereas pixels were extracted from previous frames, , is composed of pixels. For this reason, is now replaced by in Table 1. In addition, the number of extracted pixels is referred as to since they could be different from .
- Once the first hyperspectral frames were processed in order to obtain the background statistic needed by the anomaly detection process, the set of core operations are applied in the next received image blocks, , in order to obtain the reference pixels needed by the compression process. As it was explained in Section 2.2.3, these pixels are also used to detect those image blocks where anomalous pixels are present. For doing this, the Gram–Schmidt method is used to estimate the projection separation statistic, defined in Equation (3), for each pixel . If no pixels within are anomalous, other pixels within are not analyzed. Otherwise, the entire image block, , is processed to also detect mixed anomalous pixels. The number of FLOPs involved in this last step are collected in the last two rows of Table 1. The first one collects the number of FLOPs for those free of anomalous while the second one represents the opposite case. It should be emphasized that this last situation is very unlikely since anomalous pixels have normally a low presence.
3. Results
3.1. Reference Hyperspectral Data
3.2. Evaluation Metrics
3.3. Evaluation of the Proposed Method
3.3.1. Compression Performance
3.3.2. Anomaly Detection Performance
3.3.3. Computational Complexity
4. Discussions
- Line-by-Line performance. Since blocks of image pixels can be independently processed without any alignment required, it avoids accumulating high amount of uncompressed data. Hence, it permits reducing the amount of data to be processed and transferred. As a consequence, the proposed method becomes an ideal solution for applications under tight latency constraints or with limited available resources, such as memory, power and computational capabilities. In addition, the proposed algorithm fulfills the requirements imposed by applications based on push-broom/whisk-broom scanners, which permits commencement of the compression and anomaly detection processes as soon as a block of pixels is sensed.
- Low computational complexity and high level of parallelism of the most computationally demanding operations of the algorithm. This eases the hardware implementation of the proposal and reduces the amount of required hardware resources.
- Reduction in the employed hardware resources. Since the proposed methodology is based on a set of core operations common to several processes, the amount of hardware resources needed for their execution are considerably less than if different state-of-the-art algorithms were independently implemented.
- Accurate detection performance. The detection results obtained by the proposed method, which is executed online together with the compression process, are exactly the same as those provided when both the lossy compression and anomaly detection processes are executed as two independent algorithms.
5. Conclusions
Author Contributions
Funding
Conflicts of Interest
References
- Goetz, A.F. Three decades of hyperspectral remote sensing of the Earth: A personal view. Remote Sens. Environ. 2009, 113, S5–S16. [Google Scholar] [CrossRef]
- Richards, J.A. Remote Sensing Digital Image Analysis; Springer: Berlin/Heidelberg, Germany, 2013. [Google Scholar]
- Birk, A.; Wiggerich, B.; Bülow, H.; Pfingsthorn, M.; Schwertfeger, S. Safety, security, and rescue missions with an unmanned aerial vehicle (UAV). J. Intell. Robot. Syst. 2011, 64, 57–76. [Google Scholar] [CrossRef]
- Transon, J.; d’Andrimont, R.; Maugnard, A.; Defourny, P. Survey of hyperspectral Earth Observation applications from space in the Sentinel-2 context. Remote Sens. 2018, 10, 157. [Google Scholar] [CrossRef] [Green Version]
- Govender, M.; Chetty, K.; Bulcock, H. A review of hyperspectral remote sensing and its application in vegetation and water resource studies. Water Sa 2007, 33. [Google Scholar] [CrossRef] [Green Version]
- Ghamisi, P.; Yokoya, N.; Li, J.; Liao, W.; Liu, S.; Plaza, J.; Rasti, B.; Plaza, A. Advances in Hyperspectral Image and Signal Processing: A Comprehensive Overview of the State of the Art. IEEE Geosci. Remote Sens. Mag. 2017, 5, 37–78. [Google Scholar] [CrossRef] [Green Version]
- Lu, G.; Fei, B. Medical hyperspectral imaging: A review. J. Biomed. Opt. 2014, 19, 010901. [Google Scholar] [CrossRef]
- Khan, M.J.; Khan, H.S.; Yousaf, A.; Khurshid, K.; Abbas, A. Modern trends in hyperspectral image analysis: A review. IEEE Access 2018, 6, 14118–14129. [Google Scholar] [CrossRef]
- Horstrand, P.; Guerra, R.; Rodríguez, A.; Díaz, M.; López, S.; López, J.F. A UAV platform based on a hyperspectral sensor for image capturing and on-board processing. IEEE Access 2019, 7, 66919–66938. [Google Scholar] [CrossRef]
- Plaza, A.; Benediktsson, J.A.; Boardman, J.W.; Brazile, J.; Bruzzone, L.; Camps-Valls, G.; Chanussot, J.; Fauvel, M.; Gamba, P.; Gualtieri, A.; et al. Recent advances in techniques for hyperspectral image processing. Remote Sens. Environ. 2009, 113, S110–S122. [Google Scholar] [CrossRef]
- Ortiz, A.; Rodríguez, A.; Guerra, R.; López, S.; Otero, A.; Sarmiento, R.; De la Torre, E. A runtime-scalable and hardware-accelerated approach to on-board linear unmixing of hyperspectral images. Remote Sens. 2018, 10, 1790. [Google Scholar] [CrossRef] [Green Version]
- Villafranca, A.G.; Corbera, J.; Martín, F.; Marchán, J.F. Limitations of hyperspectral earth observation on small satellites. J. Small Satell. 2012, 1, 19–29. [Google Scholar]
- Valentino, R.; Jung, W.S.; Ko, Y.B. A Design and Simulation of the Opportunistic Computation Offloading with Learning-Based Prediction for Unmanned Aerial Vehicle (UAV) Clustering Networks. Sensors 2018, 18, 3751. [Google Scholar] [CrossRef] [Green Version]
- Lopez, S.; Vladimirova, T.; Gonzalez, C.; Resano, J.; Mozos, D.; Plaza, A. The promise of reconfigurable computing for hyperspectral imaging onboard systems: A review and trends. Proc. IEEE 2013, 101, 698–722. [Google Scholar] [CrossRef]
- George, A.D.; Wilson, C.M. Onboard processing with hybrid and reconfigurable computing on small satellites. Proc. IEEE 2018, 106, 458–470. [Google Scholar] [CrossRef]
- Bioucas-Dias, J.M.; Plaza, A.; Camps-Valls, G.; Scheunders, P.; Nasrabadi, N.; Chanussot, J. Hyperspectral remote sensing data analysis and future challenges. IEEE Geosci. Remote Sens. Mag. 2013, 1, 6–36. [Google Scholar] [CrossRef] [Green Version]
- Lu, K.; Xie, J.; Wan, Y.; Fu, S. Toward UAV-Based Airborne Computing. IEEE Wirel. Commun. 2019, 26, 172–179. [Google Scholar] [CrossRef]
- Interuniversity Microelectronics Centre (IMEC). Hyperspectral Drone Camera System for Application Development. Available online: https://www.imec-int.com/drupal/sites/default/files/inline-files/UAV%20SNmosaic%20VIS%2BNIR%20hyperspectral%20imaging%20camera.pdf (accessed on 13 April 2020).
- Fu, S.; Chang, R.; Couture, S.; Menarini, M.; Escobar, M.; Kuteifan, M.; Lubarda, M.; Gabay, D.; Lomakin, V. Micromagnetics on high-performance workstation and mobile computational platforms. J. Appl. Phys. 2015, 117, 17E517. [Google Scholar] [CrossRef]
- Díaz, M.; Guerra, R.; Horstrand, P.; Martel, E.; López, S.; López, J.F.; Roberto, S. Real-Time Hyperspectral Image Compression Onto Embedded GPUs. IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens. 2019, 12, 2792–2809. [Google Scholar] [CrossRef]
- Plaza, A.; Du, Q.; Chang, Y.L.; King, R.L. High performance computing for hyperspectral remote sensing. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2011, 4, 528–544. [Google Scholar] [CrossRef]
- Alcolea, A.; Paoletti, M.E.; Haut, J.M.; Resano, J.; Plaza, A. Inference in Supervised Spectral Classifiers for On-Board Hyperspectral Imaging: An Overview. Remote Sens. 2020, 12, 534. [Google Scholar] [CrossRef] [Green Version]
- Du, Q.; Nekovei, R. Fast real-time onboard processing of hyperspectral imagery for detection and classification. J. Real-Time Image Process. 2009, 4, 273–286. [Google Scholar] [CrossRef]
- Lentaris, G.; Maragos, K.; Stratakos, I.; Papadopoulos, L.; Papanikolaou, O.; Soudris, D.; Lourakis, M.; Zabulis, X.; Gonzalez-Arjona, D.; Furano, G. High-Performance Embedded Computing in Space: Evaluation of Platforms for Vision-Based Navigation. J. Aerosp. Inf. Syst. 2018, 15, 178–192. [Google Scholar] [CrossRef]
- Orlandić, M.; Fjeldtvedt, J.; Johansen, T.A. A Parallel FPGA Implementation of the CCSDS-123 Compression Algorithm. Remote Sens. 2019, 11, 673. [Google Scholar] [CrossRef] [Green Version]
- Plaza, A.J. Special issue on architectures and techniques for real-time processing of remotely sensed images. J. Real-Time Image Process. 2009, 4, 191–193. [Google Scholar] [CrossRef]
- Ratle, F.; Camps-Valls, G.; Weston, J. Semisupervised neural networks for efficient hyperspectral image classification. IEEE Trans. Geosci. Remote Sens. 2010, 48, 2271–2282. [Google Scholar] [CrossRef]
- Christophe, E. Hyperspectral data compression tradeoff. In Optical Remote Sensing; Springer: Cham, Switzerland, 2011; pp. 9–29. [Google Scholar]
- Hussain, A.J.; Al-Fayadh, A.; Radi, N. Image compression techniques: A survey in lossless and lossy algorithms. Neurocomputing 2018, 300, 44–69. [Google Scholar] [CrossRef]
- Motta, G.; Rizzo, F.; Storer, J.A. Hyperspectral Data Compression; Springer Science & Business Media: Cham, Switzerland, 2006. [Google Scholar]
- Horstrand, P.; López, J.F.; López, S.; Leppälampi, T.; Pusenius, M.; Rooker, M. A Simulation Environment for Validation and Verification of Real Time Hyperspectral Processing Algorithms on-Board a UAV. Remote Sens. 2019, 11, 1852. [Google Scholar] [CrossRef] [Green Version]
- Su, H.; Du, P.; Du, Q. Semi-supervised dimensionality reduction using orthogonal projection divergence-based clustering for hyperspectral imagery. Opt. Eng. 2012, 51, 111715. [Google Scholar] [CrossRef]
- Bioucas-Dias, J.M.; Plaza, A.; Dobigeon, N.; Parente, M.; Du, Q.; Gader, P.; Chanussot, J. Hyperspectral Unmixing Overview: Geometrical, Statistical, and Sparse Regression-Based Approaches. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2012, 5, 354–379. [Google Scholar] [CrossRef] [Green Version]
- Chang, C.; Xiong, W.; Chen, H.; Chai, J. Maximum Orthogonal Subspace Projection Approach to Estimating the Number of Spectral Signal Sources in Hyperspectral Imagery. IEEE J. Sel. Top. Signal Process. 2011, 5, 504–520. [Google Scholar] [CrossRef] [Green Version]
- Bernabe, S.; Lopez, S.; Plaza, A.; Sarmiento, R.; Rodriguez, P.G. FPGA Design of an Automatic Target Generation Process for Hyperspectral Image Analysis. In Proceedings of the IEEE 17th International Conference on Parallel and Distributed Systems, Tainan, Taiwan, 7–9 December 2011; pp. 1010–1015. [Google Scholar]
- Li, H.; Chang, C. Linear spectral unmixing using least squares error, orthogonal projection and simplex volume for hyperspectral images. In Proceedings of the 7th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS), Tokyo, Japan, 2–5 June 2015; pp. 1–4. [Google Scholar] [CrossRef]
- Kwon, H.; Nasrabadi, N.M. Kernel orthogonal subspace projection for hyperspectral signal classification. IEEE Trans. Geosci. Remote Sens. 2005, 43, 2952–2962. [Google Scholar] [CrossRef]
- Ren, H.; Chang, C.-I. Automatic spectral target recognition in hyperspectral imagery. IEEE Trans. Aerosp. Electron. Syst. 2003, 39, 1232–1249. [Google Scholar] [CrossRef] [Green Version]
- Saad, Y. Numerical Methods for Large Eigenvalue Problems; Manchester University Press: Manchester, UK, 1992. [Google Scholar]
- Guerra, R.; Santos, L.; López, S.; Sarmiento, R. A new fast algorithm for linearly unmixing hyperspectral images. IEEE Trans. Geosci. Remote Sens. 2015, 53, 6752–6765. [Google Scholar] [CrossRef]
- Guerra, R.; Barrios, Y.; Díaz, M.; Santos, L.; López, S.; Sarmiento, R. A New Algorithm for the On-Board Compression of Hyperspectral Images. Remote Sens. 2018, 10, 428. [Google Scholar] [CrossRef] [Green Version]
- Díaz, M.; Guerra, R.; López, S.; Sarmiento, R. An algorithm for an accurate detection of anomalies in hyperspectral images with a low computational complexity. IEEE Trans. Geosci. Remote Sens. 2017, 56, 1159–1176. [Google Scholar] [CrossRef]
- Díaz, M.; Guerra, R.; Horstrand, P.; López, S.; Sarmiento, R. A Line-by-Line Fast Anomaly Detector for Hyperspectral Imagery. IEEE Trans. Geosci. Remote Sens. 2019, 57, 8968–8982. [Google Scholar] [CrossRef]
- Guerra, R.; Barrios, Y.; Díaz, M.; Baez, A.; López, S.; Roberto, S. A Hardware-Friendly Hyperspectral Lossy Compressor for Next-Generation Space-Grade Field Programmable Gate Arrays. IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens. 2019, 12, 4813–4828. [Google Scholar] [CrossRef]
- Sánchez, S.; Ramalho, R.; Sousa, L.; Plaza, A. Real-time implementation of remotely sensed hyperspectral image unmixing on GPUs. J. Real-Time Image Process. 2015, 10, 469–483. [Google Scholar] [CrossRef]
- Gonzalez, C.; Sánchez, S.; Paz, A.; Resano, J.; Mozos, D.; Plaza, A. Use of FPGA or GPU-based architectures for remotely sensed hyperspectral image processing. Integration 2013, 46, 89–103. [Google Scholar] [CrossRef]
- Horstrand, P.; Diaz, M.; Guerra, R.; Lopez, S.; Lopez, J.F. A Novel Hyperspectral Anomaly Detection Algorithm for Real-Time Applications With Push-Broom Sensors. IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens. 2019, 12, 4787–4797. [Google Scholar] [CrossRef]
- Diaz, M.; Guerra Hernández, R.; Lopez, S. A Novel Hyperspectral Target Detection Algorithm For Real-Time Applications With Push-Broom Scanners. In Proceedings of the 10th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS), Amsterdam, The Netherlands, 24–26 September 2019; pp. 1–5. [Google Scholar] [CrossRef]
- Nasrabadi, N.M. Hyperspectral target detection: An overview of current and future challenges. IEEE Signal Process. Mag. 2014, 31, 34–44. [Google Scholar] [CrossRef]
- Consultative Committee for Space Data Systems (CCSDS). Blue Books: Recommended Standards. Available online: https://public.ccsds.org/Publications/BlueBooks.aspx (accessed on 23 March 2019).
- Howard, P.G.; Vitter, J.S. Fast and efficient lossless image compression. In Proceedings of the Data Compression Conference, DCC’93, IEEE, Snowbird, UT, USA, 30 March–2 April 1993; pp. 351–360. [Google Scholar]
- Díaz, M.; Guerra, R.; López, S. A Hardware-Friendly Anomaly Detector for Real-Time Applications With Push-Broom Scanners. In Proceedings of the 10th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS), Amsterdam, The Netherlands, 24–26 September 2019; pp. 1–5. [Google Scholar] [CrossRef]
- Specim FX1 Series Hyperspectral Cameras. Available online: http://www.specim.fi/fx/ (accessed on 23 March 2019).
- DJI, MATRICE 600 PRO. Available online: https://www.dji.com/bg/matrice600 (accessed on 23 March 2019).
- Guerra, R.; Horstrand, P.; Rodríguez, A.; Díaz, M.; Morales, A.; Jiménez, A.; López, S.; López, J.F. Optimal UAV movement control for farming area scanning using hyperspectral pushbroom sensors. In Proceedings of the XXXIV Conference on Design of Circuits and Integrated Systems (DCIS), IEEE, Bilbao, Spain, 20–22 November 2019; pp. 1–6. [Google Scholar]
- Wang, Z.; Bovik, A.C.; Sheikh, H.R.; Simoncelli, E.P. Image quality assessment: From error visibility to structural similarity. IEEE Trans. Image Process. 2004, 13, 600–612. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Jafarzadeh, H.; Hasanlou, M. An Unsupervised Binary and Multiple Change Detection Approach for Hyperspectral Imagery Based on Spectral Unmixing. IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens. 2019, 12, 4888–4906. [Google Scholar] [CrossRef]
- Jafarzadeh, H.; Hasanlou, M. Assessing and comparing the performance of endmember extraction methods in multiple change detection using hyperspectral data. In Proceedings of the International Archives of the Photogrammetry, Remote Sensing & Spatial Information Sciences, Karaj, Iran, 12–14 October 2019. [Google Scholar]
- Chang, C.I.; Chiang, S.S. Anomaly detection and classification for hyperspectral imagery. IEEE Trans. Geosci. Remote Sens. 2002, 40, 1314–1325. [Google Scholar] [CrossRef] [Green Version]
- Zhao, C.; Deng, W.; Yan, Y.; Yao, X. Progressive line processing of kernel RX anomaly detection algorithm for hyperspectral imagery. Sensors 2017, 17, 1815. [Google Scholar] [CrossRef] [Green Version]
- Aiazzi, B.; Alparone, L.; Baronti, S. Quality issues for compression of hyperspectral imagery through spectrally adaptive DPCM. In Satellite Data Compression; Springer: Cham, Switzerland, 2012; pp. 115–147. [Google Scholar]
- Lee, C.; Lee, S.; Lee, J. Effects of lossy compression on hyperspectral classification. In Satellite Data Compression; Springer: Cham, Switzerland, 2012; pp. 269–285. [Google Scholar]
- Garcia-Vilchez, F.; Muñoz-Marí, J.; Zortea, M.; Blanes, I.; González-Ruiz, V.; Camps-Valls, G.; Plaza, A.; Serra-Sagristà, J. On the impact of lossy compression on hyperspectral image classification and unmixing. IEEE Geosci. Remote Sens. Lett. 2010, 8, 253–257. [Google Scholar] [CrossRef]
- Du, Q.; Ly, N.; Fowler, J.E. An operational approach to PCA+ JPEG2000 compression of hyperspectral imagery. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2013, 7, 2237–2245. [Google Scholar] [CrossRef]
- Chang, C.I. Hyperspectral Data Processing: Algorithm Design and Analysis; John Wiley & Sons: New York, NY, USA, 2013. [Google Scholar]
Stage | FLOPs | Complexity | |
---|---|---|---|
Core operations | |||
Codification | |||
Core operations (AD) | |||
Core operations | |||
Codification | |||
Core operations | |||
Codification | |||
Gram–Schmidt | |||
(non-anomalous) | |||
Gram–Schmidt | |||
(anomalous) |
Inputs | Image 1 | Image 2 | Image 3 | Image 4 | Image 5 | Image 6 | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Nbits | BS | CR | CR | bpppb | CR | bpppb | CR | bpppb | CR | bpppb | CR | bpppb | CR | bpppb |
12 | 1024 | 12 | 15.87 | 1.01 | 15.99 | 1.00 | 15.97 | 1.00 | 17.05 | 0.94 | 16.13 | 0.99 | 15.73 | 1.02 |
16 | 22.56 | 0.71 | 22.81 | 0.70 | 22.82 | 0.70 | 24.43 | 0.65 | 22.83 | 0.70 | 22.16 | 0.72 | ||
20 | 28.47 | 0.56 | 28.88 | 0.55 | 28.98 | 0.55 | 31.05 | 0.56 | 28.75 | 0.55 | 27.78 | 0.58 |
Inputs | Image 1 | Image 2 | Image 3 | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Nbits | BS | CR | SNR | MAD | SSIM | RMSE | PSNR | SNR | MAD | SSIM | RMSE | PSNR | SNR | MAD | SSIM | RMSE | PSNR |
12 | 1024 | 12 | 45.87 | 1235.00 | 0.998 | 118.89 | 54.83 | 45.53 | 1318.00 | 0.998 | 110.46 | 55.46 | 45.72 | 946.00 | 0.998 | 96.43 | 56.65 |
16 | 43.88 | 1504.00 | 0.998 | 149.46 | 52.84 | 43.76 | 1621.00 | 0.998 | 135.52 | 53.69 | 44.39 | 1170.00 | 0.998 | 112.42 | 55.31 | ||
20 | 42.66 | 2245.00 | 0.998 | 171.96 | 51.62 | 42.67 | 1817.00 | 0.997 | 153.62 | 52.60 | 43.54 | 1343.00 | 0.997 | 123.93 | 54.47 |
Inputs | Image 4 | Image 5 | Image 6 | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Nbits | BS | CR | SNR | MAD | SSIM | RMSE | PSNR | SNR | MAD | SSIM | RMSE | PSNR | SNR | MAD | SSIM | RMSE | PSNR |
12 | 1024 | 12 | 34.99 | 1340.00 | 0.997 | 121.11 | 54.67 | 35.43 | 2429.00 | 0.990 | 239.26 | 48.75 | 34.79 | 2936.00 | 0.988 | 281.37 | 47.34 |
16 | 34.40 | 1557.00 | 0.997 | 129.54 | 54.08 | 34.84 | 2682.00 | 0.989 | 256.09 | 48.16 | 34.06 | 3496.00 | 0.987 | 305.97 | 46.62 | ||
20 | 34.00 | 1694.00 | 0.997 | 135.75 | 53.67 | 34.43 | 2916.00 | 0.988 | 268.47 | 47.75 | 33.55 | 3799.00 | 0.985 | 324.68 | 46.10 |
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Díaz, M.; Guerra, R.; Horstrand, P.; López, S.; López, J.F.; Sarmiento, R. Towards the Concurrent Execution of Multiple Hyperspectral Imaging Applications by Means of Computationally Simple Operations. Remote Sens. 2020, 12, 1343. https://doi.org/10.3390/rs12081343
Díaz M, Guerra R, Horstrand P, López S, López JF, Sarmiento R. Towards the Concurrent Execution of Multiple Hyperspectral Imaging Applications by Means of Computationally Simple Operations. Remote Sensing. 2020; 12(8):1343. https://doi.org/10.3390/rs12081343
Chicago/Turabian StyleDíaz, María, Raúl Guerra, Pablo Horstrand, Sebastián López, José F. López, and Roberto Sarmiento. 2020. "Towards the Concurrent Execution of Multiple Hyperspectral Imaging Applications by Means of Computationally Simple Operations" Remote Sensing 12, no. 8: 1343. https://doi.org/10.3390/rs12081343