Figure 4 - available via license: Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
Content may be subject to copyright.
Subset of the result of integrated model for the (a) original Quickbird image and (b) pan-sharpened image (Own Analysis, 2016)
Source publication
Image-sharpening process integrates lower spatial resolution multispectral bands with higher spatial resolution panchromatic band to produce multispectral bands with finer spatial detail called pan-sharpened image. Although the pan-sharpened image can greatly assist the process of information extraction using visual interpretation, the benefit and...
Context in source publication
Context 1
... inputs were used to produce IM bands using formula described in Wicaksono (2010). The result of the integrated model is shown in Figure 4. There are not many differences in the IM bands of the original and pan-sharpened image, except the amount of sunglint is greater on pan- sharpened image. ...
Similar publications
Multi-temporal change detection over decades including the pre-satellite era is challenging due to the different image types available over time, and this explains the scarcity of long-term studies of vegetation succession which can play a pivotal role in the restoration of biodiversity in regenerating forests. This study describes a semi-automated...
Citations
... The pixel-based classification improved overall accuracy (47.57%) in mapping benthic habitats compared with the object-based classification approach (36.17%). In the Kemujan Island study using Quickbird satellite imagery, the overall accuracy levels of pixel-based image classification and the ML method were 64.28% and 73.30%, respectively (Wicaksono et al., 2017). ...
... (Tamondong et al., 2013), and yield greater accuracy than medium-resolution photo types such as IKONOS (OA = 80%) (Tran Van, 2008), Quickbird satellite images (OA = 73.30%) (Wicaksono et al., 2017), and ALOS AVNIR-2 (OA = 78%) (Wicaksono, Fauzan, et al., 2019). However, classification based on remotesensing images for areas with low seagrass density is complex. ...
Benthic habitats are critical in shallow sea areas; they regulate the diversity and richness of organisms in each area. Mapping benthic habitats elucidates natural sea characteristics and aids in managing and using natural resources, as well as conserving marine biodiversity. This study established a benthic habitat map for the Da Lon Reef area, Truong Sa Islands, Vietnam, using Pléiades high-resolution remote sensing imaging materials and field survey results from 2020 and 2021. We identified seven classes of benthic habitats with a 91.64% overall accuracy, corresponding to a Kappa coefficient of 0.88. In the Da Lon Reef, seagrass biomes occupy a large area (more than 200 ha) and are distributed mainly inside lagoons at depths of 2–6 m. The field survey results identified five seagrass species and the biodiversity and biomass of seagrass populations in the lagoon of Da Lon Reef. The study results confirm the fundamental value of resources, biodiversity in general and seagrass in particular, in managing and protecting shallow sea ecosystems and biodiversity conservation in the Da Lon Reef area, an important part of the Truong Sa Islands, Vietnam.
... [44]. For benthic habitat mapping study, GEOBIA usage is produced higher accuracy (73.3% while using sun glint corrected image) compared to the per-pixel classification (64.3%) and is comparable to the result of the original image [60]. Wahidin et al [61] had tried classified Landsat 8 OLI image coral for reef benthic habitat mapping employing OBIA within seven classes (sand-rubble, rubble, coral, sand, sand-algae, sea-grass, and rubble-sand). ...
The traditionally human-visual or manual interpretation is generally chosen method for spatial feature extraction on map producing process in Indonesia, although the pixel-based feature extraction is used in particular step. The main problem of human-visual method is the dependency to operator capacity. Subjectivity may occur because each operator has different ability. In early 20th century, the image segmentation method was still developing and now Geographic Object-based Image Analysis (GEOBIA) method is introduced to mapping process. Researchers are still developing and applying GEOBIA for many mapping purposes. The advantage of GEOBIA is object-based segmentation which contextually influenced by spatial feature characteristics. GEOBIA provides more meaningful information than traditional pixel-based image analysis by allowing less well-defined edges or borders between different classes. On the resulting maps, there are divisions between different types of object classification, for example where some species of mangroves meet are not generally represented by a single object. The developing of remotely sensed imagery data quality requires the developing of extraction method as well. This article discusses the implementation of GEOBIA for coastal mapping in Indonesia. Some research combined pixel-based and GEOBIA method, and some get the overall accuracy > 70% of GEOBIA method. By developing a new segmentation accuracy measure, we evaluated that segmentation accuracies decrease with increasing segmentation scales and the negative impact of under-segmentation errors become significantly rise at large scale. The use of GEOBIA for coastal mapping in Indonesia are still lacking, locally applicable and still developing the transferability rule sets. The conclusion is that the innovation is needed to make GEOBIA method applicable for mass production of Indonesian coastal map.
The adoption of remotely sensed data in forest applications has grown significantly. Whereas high spatial resolution sensors have been successful in mapping and monitoring commercial forests, their cost, accessibility, and spatial coverage remain a critical challenge. Hence, it is was necessary to investigate the value of new and improved freely available sensors in forest species mapping using the Partial Least Square-Discriminant Analysis (PLS-DA). This study evaluated the performance of new freely available and improved raw and pan-sharpened Landsat 8 Operational Land Imager (OLI) imagery in discriminating seven key plantation forest species in KwaZulu-Natal, South Africa. Accuracies achieved using the Landsat (OLI) imagery were benchmarked against the WorldView-2 imagery. Results show that raw and pan-sharpened bands successfully delineated commercial forest species, with overall classification accuracies of 79% and 77.8%, respectively. Although these accuracies were lower than the 86.5% achieved from the higher resolution Worldview-2 image data, our findings demonstrate that the Landsat 8 OLI’s lower spatial resolution (30 m) generated a plausible performance in discriminating forest species. Hence, Landsat 8 OLI could be useful in providing existing and historical preliminary forestry assessment due to its free availability, wide spatial coverage as well as its rich archive dating back to the 1970s.
