| 585 | = Generation of reference data = |
| 586 | |
| 587 | High resolution reference data were necessary for this study, in order to train and validate the classi-fication algorithms. |
| 588 | However, the only available European-wide data, was the JRC Forest Cover Map, compiled in 2006. Since our datasets were |
| 589 | collected in 2009, the Landsat TM data, acquired in 2009, were classified in order to produce a stocked/non-stocked classification |
| 590 | map, which would be used to evaluate the classification algorithms in the medium/low resolution datasets. |
| 591 | = Generation of reference data = |
| 592 | |
| 593 | High resolution reference data were necessary for this study, in order to train and validate the classi-fication algorithms. |
| 594 | However, the only available European-wide data, was the JRC Forest Cover Map, compiled in 2006. Since our datasets were collected |
| 595 | in 2009, the Landsat TM data, acquired in 2009, were classified in order to produce a stocked/non-stocked classification map, |
| 596 | which would be used to evaluate the classification algorithms in the medium/low resolution datasets. |
| 597 | |
| 598 | For the classification of Landsat TM in stocked and non-stocked areas, a set of 400 points was generated by stratified random |
| 599 | sampling, following the estimation of the extent of the two classes through an unsupervised classification. These points were |
| 600 | used as training samples, and were identified as stocked or non-stocked by using the JRC Forest Cover Map and visual interpretation |
| 601 | of Google Earth images. (Figure 2). A radius of two pixels was considered around each point in order to identify it as stocked |
| 602 | or non-stocked. A second set of 300 points was generated using the same method, to be used for the validation of the classification. |
| 603 | A Maximum Likelihood supervised classification was performed on three different band combina-tions: 6 bands (excluding the thermal band), |
| 604 | 4 bands (Blue, Red, NIR and SWIR), and 4 bands + NDVI generated from the Landsat TM data, in order to find the best result. The same |
| 605 | training samples were used for the three band combinations. The four classes defined in these classifications were "stocked areas", |
| 606 | "non-stocked areas", “water” and "No data". The confusion matrices, for the three classification’s results, were performed using |
| 607 | the same 300 validation points. |
| 608 | = Forest cover reference maps creation = |
| 609 | |
| 610 | |
| 611 | The classification map with the higher accuracy was resampled to a 10 m resolution, using nearest neighbour,and the “no data” |
| 612 | and “water” classes were masked out, leaving the two main classes, “stocked” and “non-stocked” with values of 1 and 0, respectively. |
| 613 | |
| 614 | In order to create reference maps with the same resolution as the medium/low resolution images, the 10 m resolution map was |
| 615 | aggregated to 250, 300 and 1000 m resolution. The new value of each aggregated pixel was the percentage of the original “stocked” |
| 616 | pixels that were aggregated to the lower resolution pixel. According to the pixel values obtained after aggregation, each map was |
| 617 | reclassified to five classes according to the percentage of forest cover (0 – 10%, 11% - 30%, 31% - 50%, 51% - 75%, 76% - 100%), |
| 618 | see Figure 3. Those classes were chosen in order to ensure that the entire range of forest cover fractions would be |
| 619 | represented within the broad “stocked” and “non-stocked” classes. |
| 620 | |
| 621 | The Maximum Likelihood supervised classification on the three different combinations of Landsat TM bands produced the highest |
| 622 | overall accuracy (93.12%) and Kappa Coefficient (0.904) when all six bands (excluding the thermal) were used, as shown in Table 2. |
| 623 | The resulting stocked/non-stocked map was used as a reference map for the study. |
| 624 | |
| 625 | {| border="1" |
| 626 | |+ Results of the accuracy assessment of Landsat TM classification |
| 627 | ! !! TM 6 bands !!TM 4 bands !! TM4 + NDVI |
| 628 | |- |
| 629 | ! Overall accuracy |
| 630 | | 93.1214% || 92.9878% || 92.4135% |
| 631 | |- |
| 632 | ! Kappa coefficient |
| 633 | | 0.9045 || 0.9027 || 0.8947 |
| 634 | |} |
| 635 | |
| 636 | = Evaluation of classification algorithms = |
| 637 | |
| 638 | |
| 639 | The produced reference data were used to generate training samples for the classifications and the validation points |
| 640 | for the confusion matrices. Training samples were generated for the five clas-ses using stratified random sampling method and |
| 641 | subsequently the regions of interests were merged in order to create the two main classes, Non-stocked (0% - 10% and 11% - 30%) |
| 642 | and Stocked (31% - 50%, 51% - 75%, and 76% - 100%).These training samples were used to perform the Maximum Likelihood (ML), |
| 643 | Support Vector Machine (SVM) and Artificial Neural Network (ANN) supervised classifications on the two- and three-band MODIS, |
| 644 | VGT and PROBA-V datasets. Con-fusion matrices were generated for all classification results using the LandsatTM-derived 250m, |
| 645 | 300m and 1000m reference maps , respectively for MODIS, PROBA-V and VGT classifications. |
| 646 | = Results = |
| 647 | |
| 648 | |
| 649 | RESULTS |
| 650 | The results of the confusion matrices are presented in Figure 4. It appeared that the two-band 250 m MODIS |
| 651 | dataset produced more accurate classifications with all classification methods, in comparison with the |
| 652 | three-band dataset, which employed the SWIR data resampled from 500 to 250 metres. |
| 653 | Comparison between the classification methods for all datasets shows that the ANN classification gives a slightly higher |
| 654 | overall accuracy than the other methods for the two- and three-band MODIS data, as well as the simulated PROBA-V, while |
| 655 | SVM gives the highest overall accuracy for VGT. However the differences between the different classifiers for each dataset |
| 656 | are very small and could be considered to be of low significance. |
| 657 | According to these preliminary results, the best accuracy and kappa coefficients were achieved by the ANN classification on |
| 658 | MODIS image with two bands at 250 m spatial resolution and the lowest overall accuracy was performed by ML classification on |
| 659 | simulated PROBA-V image . |
| 660 | |
| 661 | In addition to the classification accuracy the percentage of stocked area distribution in each map was also calculated . |
| 662 | For MODIS and VGT data, the estimated forest cover was overestimated significantly, from 27% which was the estimated |
| 663 | forest cover using the Landsat TM data, to 41 to 48%. On the contrary forest cover estimation with PROBA-V, using the |
| 664 | small subset of the scene, for which the simulated PROBA-V data were available, showed significant differ-ences between |
| 665 | the classifiers. The Landsat TM data produced an approximate 70% forest cover, while Maximum Likelihood underestimated the |
| 666 | forest cover (57.55%) and MODIS and VGT over-estimated the forest cover (80.72 and 79.39% respectively). |
| 667 | |
| 668 | {| align="center" |
| 669 | |+'''Figure 4: Overall accuracy and Kappa Coefficients or ML, ANN and SVM classification result on MODIS, VGT and PROBA-V (B: bands)''' |
| 670 | |- |
| 671 | |[wiki:File:Table.png File:Table.png] |
| 672 | |} |
| 673 | |
| 674 | |
| 675 | {| align="center" |
| 676 | |+'''Figure 5: Class distribution of the different classification results comparing to Landsat TM refer-ence (B: bands).''' |
| 677 | |- |
| 678 | |[wiki:File:Table2.png File:Table2.png] |
| 679 | |} |
| 680 | |
| 681 | |
| 682 | {| align="center" |
| 683 | |+'''Figure 6: Classes distribution of simulated PROBA-V classification compared to the subset of the Landsat reference map.''' |
| 684 | |- |
| 685 | |[wiki:File:Table3.png File:Table3.png] |
| 686 | |} |
| 687 | = Discussion and Conclusions = |
| 688 | |
| 689 | |
| 690 | The error matrix is the most common method to evaluate classification accuracy (4), and it is the starting point |
| 691 | for many analysis techniques (4,5). In addition to the overall accuracy, the entire confusion matrix was considered |
| 692 | through the use of the kappa statistic. These analyses were used to allow a comparison between classification results. |
| 693 | |
| 694 | Forest cover in the study area showed significant fragmentation, which tends to pose problems when classifying satellite |
| 695 | images (6). This fact proved to be detrimental to the performance of cer-tain datasets. The main aim of this study was to |
| 696 | compare the potential performance of PROBA-V in estimating forest cover, with that of VEGETATION (VGT). Because of the small |
| 697 | extent of the simulated PROBA-V scene, 250 m MODIS data were also evaluated as an alternative sensor with comparable spatial |
| 698 | resolution, which could deliver more statistically significant results. The lack of 250 m SWIR data (the native 250 m product |
| 699 | contains data only in the Red and NIR channels) forced the resampling of the 500 m SWIR data to 250 m. Comparison between the |
| 700 | classification accuracies between the three-band (with the resampled SWIR data) and the two-band (without the SWIR data) |
| 701 | MODIS data showed that the latter achieves more accurate classifications. Spectral variation of the SWIR signal within the |
| 702 | 500 m pixel could not be retrieved with the resampling pro-cess and, as a result, the combination of accurate 250 m Red and |
| 703 | NIR data with “false” SWIR data led to considerable misclassifications between stocked and non-stocked areas On the contra-ry, |
| 704 | the use of just the Red and NIR data produced more accurate classifications. |
| 705 | |
| 706 | Comparison between classifications of the two-band MODIS data and the VGT data, revealed that the former again produced |
| 707 | more accurate classifications. The availability of SWIR data in the VGT dataset could not assist the classification |
| 708 | sufficiently in order to make up for the increased uncer-tainty brought about by the 1 km spatial resolution of the data. |
| 709 | As expected, extraction of forest cover information in a fragmented landscape using low resolution data, proves to |
| 710 | be quite prob-lematic at a local to regional scale (4). |
| 711 | |
| 712 | The simulated PROBA-V data covered only a 20 „e 20 km2 area and consisted of approximately 3600 pixels. From a statistical |
| 713 | point of view it is very difficult to extract statistically significant re-sults from such a small sample size, which is the |
| 714 | reason MODIS data were also used in this study. Nevertheless, the classification accuracy of PROBA-V data was very similar to |
| 715 | the one produced by VEGETATION. On the other hand, the low kappa values of the classifications are caused by the amplification |
| 716 | of the errors in the confusion matrix, brought about the small sample size. |
| 717 | |
| 718 | The accuracies of the classifications produced by the different classification algorithms on the same dataset showed that |
| 719 | there is effectively no difference in their performance on all three da-tasets. The only exception was Maximum Likelihood |
| 720 | classifier on the simulated PROBA-V data, which was by 5% less accurate in comparison with ANN and SVM, a fact probably |
| 721 | attributed to the quality of the simulated PROBA-V data The Artificial Neural Network classification provides the most |
| 722 | accurate classifications and appears to deal best with the mixture of different cover types within each coarse resolution |
| 723 | pixel (7) and the fragmentation of the landscape (8). |
| 724 | |
| 725 | Non-thematic errors may be one major problem in the use of confusion matrix and associated ac-curacy analysis (9). This is |
| 726 | particularly true for errors of image misregistration, since the images were registered to an image of a higher spatial |
| 727 | resolution. The methodology used of aggregation and forest coverage percentage class was used to minimize these errors. |
| 728 | |
| 729 | In conclusion, comparison between classification algorithms on MODIS and VGT data, showed no significant differences between |
| 730 | them, in terms of classification accuracy. The higher accuracy achieved by those classifiers when using the MODIS data, which |
| 731 | have comparable spatial resolu-tion with PROBA-V, compared to that achieved with the VGT data, suggest that the increased spa-tial |
| 732 | resolution will provide more accurate forest cover mapping, particularly in areas with fragment-ed forest coverage. The usefulness |
| 733 | of SWIR data could not be evaluated, but it could prove useful in discriminating between forest and agricultural land, and assist |
| 734 | even further the accurate map-ping of forest cover. More research on this topic is recommended in order to quantify the useful-ness |
| 735 | of SWIR data at a 300 m resolution. |
| 736 | |
729 | | ‘occulter’ spacecraft adopting a more passive role. |
730 | | |
731 | | Following launcher separation, the stack will perform a series of perigee (or bottom of orbit) burns to raise its apogee (or top of orbit) |
732 | | to 60,524 km. This will be followed by a single apogee burn to raise its perigee to 800 km. After a short preparatory period the |
733 | | two satellites will be separated from each other and injected into a safe tandem orbit with no chance of collision. |
734 | | |
735 | | An approximately three-month commissioning period will follow. One important test will be to demonstrate that the mission’s Collision |
736 | | Avoidance Manoeuvre (CAM) functionality works. The automated CAM is activated in the event that the two satellites grow dangerously |
737 | | close, so they can be left safely in a relative orbit. |
738 | | |
739 | | |
740 | | Nominal operations will include both formation flying manoeuvres and coronagraph observation. The cost in fuel of maintaining formation |
741 | | throughout the orbit would be too great so each orbit will be divided between six hours of formation flying manoeuvres and the two hours |
742 | | closest to Earth in ‘perigee passage’, based on passive drifting with mid-course corrections. |
743 | | |
744 | | Formation flying experiments will routinely take place during the week, with coronagraph experiments being performed during the |
745 | | weekend. The coronagraph experiment will fill the entire 16 gigabit mass memory of its satellite. This data will then be progressively |
746 | | downlinked to the Redu ground station during 20 hours of passes the following week. |
747 | | == Platforms == |
748 | | |
749 | | Both the larger coronagraph and smaller occulter satellites are derived from the Proba standard platform. |
750 | | |
751 | | The 475 kg coronagraph satellite has a volume of 1010 x 1010 x 1410 mm. It hosts the coronagraph instrument which |
752 | | is pointed directly at the occulter satellite and observes only the corona of the Sun. It has a doubled H-shaped |
753 | | main structure with a set of external panels that make a box for its supporting structure on the port side of the |
754 | | satellite to hold its deployable solar array. There are also further solar cells located on the support structure itself. |
755 | | |
756 | | The satellite’s front panel allows for the opening of the coronagraph instrument, optical metrology sensors and a |
757 | | Sun sensor but no solar cells as this area will be kept in shadow by the occulter satellite during science operations. |
758 | | On the opposite side are another Sun sensor and star trackers. |
759 | | |
760 | | Its guidance, navigation and control (GNC) system consists of four reaction wheels, six gyroscopes, one star tracker |
761 | | with three heads, six Sun sensors, and two GPS systems and antennas. The majority of the formation flying system is |
762 | | located within the coronagraph satellite as it performs most of the manoeuvres. |
763 | | |
764 | | The 245 kg occulter satellite has a volume of 900 x 1100 x 900 mm. For scientific operations its task is to block |
765 | | the Sun, leaving only the solar corona visible to the coronagraph instrument. |
766 | | |
767 | | |
768 | | It is basically box-shaped, a H-shaped primary structure with all avionics and instrument equipment mounted on the |
769 | | inner panels, with an 1520-mm diameter occulting disk on the face pointing away from the Sun. The opposite side |
770 | | accommodates the satellite’s solar array. Its GNC system consists of four reaction wheels, six gyroscopes, one star |
771 | | tracker with three heads, six Sun sensors and two GPS systems and antennas. |
772 | | |
773 | | The Launch date of Proba 3 is somewhere between 2015-2016 ,expect more info [http://www.esa.int/esaMI/Proba/SEMG2R4PVFG_0.html here] |
774 | | = ESA and RTEMS validation and tools = |
775 | | |
776 | | |
777 | | Saab Space AB performed a validation of the real-time operating system RTEMS. Since it is available for many different targets and includes a multitude of functionality, ranging from I/O drivers to file-systems and beyond, it was agreed to only focus on the parts that were applicable for European space community applications. This implied that only the ERC32 target and a limited sub-set of the configurable RTEMS managers had to be considered. |
778 | | |
779 | | Subsequently, Edisoft has reached an agreement with OAR to implement an RTEMS maintenance centre (see related link) in Europe. Edisoft has complemented the validation and the toolset associated with RTEMS for the specific needs of the European space industry. Gaisler Research also provides services based around RTEMS on ERC32 and Leon. |
780 | | |
781 | | RTEMS has already been used in several space applications, in particular FedSat (a scientific Research and Development microsatellite), the Surrey's Solid State Data Recorder (a component used in the Disaster Monitoring Constellation), ChipSat (a System-on-Chip architecture), the Electra UHF antenna of the Mars Reconnaissance Orbiter and in the Galileo GIOVE-A and Herschel-Planck satellites. |
782 | | |
783 | | |
784 | | = See also = |
785 | | |
786 | | |
787 | | * [http://www.esa.int/esaMI/Proba/index.html ESA Proba Missions 400px] |
788 | | * [http://probav-iuc.org/assets/investigations/Aitkenhead.pdf Aitkenhead 400px] |
789 | | * [http://probav-iuc.org/assets/201102/PROBA-V_PDF.pdf PROBA-V 400px] |
790 | | * [http://probav-iuc.org/assets/investigations/DiBella.pdf Investigation DiBella 400px] |
791 | | * [http://www.esa.int/esaMI/Proba/SEMJJ5ZVNUF_0.html Official Page 400px] |
792 | | * [http://www.esa.int/esaMI/Proba_web_site/index.html ESA Proba1 400px] |
793 | | * [http://ilrs.gsfc.nasa.gov/docs/ESA4S_06_11d.pdf PROBA-2 MISSION AND NEW TECHNOLOGIES OVERVIEW 400px] |
794 | | * [http://ilrs.gsfc.nasa.gov/docs/AIAA_ASC_087086.pdf Autonomous and Precise Navigation of the PROBA-2 Spacecraft 400px] |
795 | | * [http://esamultimedia.esa.int/docs/Proba/Proba-2_Factsheet_8oct.pdf Factsheet 400px] |
796 | | * [http://events.eoportal.org/presentations/7111/10001905.html PROBA-V (Project for On-Board Autonomy - Vegetation) 400px] |
797 | | * [http://www.esa.int/esaCP/SEMNOJRTJRG_index_0.html NEWS about Proba2 400px] |
798 | | * [http://www.esa.int/esaCP/index.html ESA 400px] |
799 | | * [http://ilrs.gsfc.nasa.gov/docs/PROBA2_DLR_TN_0020.pdf Technical Information for PROBA-2 Laser Tracking Support 400px] |
800 | | * [[http://ilrs.gsfc.nasa.gov/satellite_missions/list_of_satellites/prb2_stadata.html Station Data]] |
801 | | * [[http://inforum.org.pt/INForum2009/docs/full/paper_88.pdf RTEMS Improvement – Space Qualification of RTEMS]] |
802 | | * [[http://news.bbc.co.uk/2/hi/science/nature/8481321.stm Esa's Proba-2 demonstration satellite views eclipse ]] |
803 | | * [[http://www.esa.int/TEC/Software_engineering_and_standardisation/TECLUMKNUQE_2.html ESA Operating Systems]] |