Standardized principal components

Singh, Ashbindu; Harrison, Andrew J.

doi:10.1080/01431168508948511

Cited by 325 publications

(119 citation statements)

References 8 publications

Supporting

Mentioning

115

Contrasting

Unclassified

Order By: Relevance

“…In the future, selection of the suitable image bands for change detection using MPCA and determination of the threshold of each bands are both important to produce highly accurate change detection results. Although the MPCA is the simplest and easiest to implement method, it proved to be too scene-dependent, which has also been reported by several authors (Singh and Harrison 1985). In addition, MPCA produce a more direct interpretation linked with the vegetation damage and recovery processes after forest fires in the study area.…”

Section: Principal Component Analysismentioning

confidence: 93%

“…The remaining component (PC6, PC7, and PC8) were quite sensitive to noise along with variation of vegetation coverage or do not display any features of significance to the study. Furthermore, some researchers found that the eigenvector characteristics of three additional principal components (PC6, PC7, and PC8) gave little more information than the first five PCs because they have eigenvalue variabilities less than 1% and did not show a useful specific form in further land cover change detection (Singh andHarrison 1985, Garcia-Haro et al 2001).…”

Section: Principal Component Analysismentioning

confidence: 99%

See 1 more Smart Citation

Detection of fire impact and vegetation recovery over tropical peat swamp forest by satellite data and ground-based NDVI instrument

Segah

Tani

Hirano

2010

International Journal of Remote Sensing

View full text Add to dashboard Cite

Multitemporal Principal Component Analysis (MPCA) was used for processing Landsat TM/ETM+ satellite images. MPCA was able to merge spectral data corresponding to TM-1996 (pre-fire in 1997), ETM-2000 (post-fire 1997 and pre-fire 2002) and ETM-2003 (post-fire in 2002), which was crucial for detecting the fire impact and vegetation recovery. Results indicate that the burnt areas of 1997 and 2002 were 89,086 ha (16.5%) and 31,859 ha (5.9%), respectively, within the study area of 540,000 ha. SPOT-VEGETATION 10-days Maximum Value Composite (MVC) data were also used and compared with Normalized Difference Vegetation Index (NDVI) from ground-based NDVI. Our research demonstrates the strong relationship between Landsat TM/ETM+, SPOT-VEGETATION data and ground-based NDVI to identifying land cover changes and vegetation recovery over the tropical peat swamp forest area in Central Kalimantan, Indonesia that is affected by forest fires that occurred in 1997 and 2002

show abstract

Section: Principal Component Analysismentioning

confidence: 93%

Section: Principal Component Analysismentioning

confidence: 99%

Detection of fire impact and vegetation recovery over tropical peat swamp forest by satellite data and ground-based NDVI instrument

Segah

Tani

Hirano

2010

International Journal of Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Most of the images obtained were captured late in the dry season, or soon after the majority of fires generally occur, to maximize mapping accuracy. The spectral signatures of burned areas could be better distinguished after principal components analysis (PCA) (Richards, 1984;Singh and Harrison, 1985;Fung and LeDrew, 1987) than in the raw data. Burned areas were subsequently separated from unburned areas with a supervised classification (parallelepiped) of the principal components (PCs).…”

Section: Fire Scar Mappingmentioning

confidence: 99%

Trends in fire patterns in a southern African savanna under alternative land use practices

Hudak

Fairbanks

Brockett

2004

Agriculture, Ecosystems & Environment

View full text Add to dashboard Cite

Climate, topography, vegetation and land use interact to influence fire regimes. Variable fire regimes may promote landscape heterogeneity, diversification in vegetation pattern and biotic diversity. The objective was to compare effects of alternative land use practices on landscape heterogeneity. Patch characteristics of fire scars were measured from 21 annual burn maps produced from 1972 to 2001 Landsat imagery. Trends in fire patterns under alternative land use practices were compared across a 250,000 ha savanna in southern Africa partitioned into three land use zones. Zone 1, Madikwe Game Reserve (MGR), has had mostly prescribed fires since 1993. Zone 2, cattle farms near MGR in South Africa (SAF), has experienced occasional fires. Zone 3, communal grazing lands in neighboring Botswana (BOT), has had the fewest fires. Cattle ranching was the predominant land use throughout the study area until 1992, when land use switched to conservation and eco-tourism in MGR. Sixteen landscape metrics were applied to this data set to uncover trends in the patch characteristics of the fire scars. A principal components analysis (PCA) reduced the dimensionality of the results so trends in the 10 most important size, shape, and proximity metrics could be better interpreted. The PCA results showed that more burning over time in MGR, and to a lesser extent in SAF, increased patch size, size variability, shape complexity and proximity, while fire exclusion in BOT produced no change or decreasing trends. We tested for significant differences in these metrics between the three land use zones and between two periods, 1972-1992 and 1993-2001. Most patch characteristics in MGR and SAF differed significantly from those in BOT, especially during the latter period, while between MGR and SAF they did not. Patch area, shape complexity and core area increased significantly between periods in MGR, while patch size, size variability and core area increased significantly between periods in SAF. In BOT, no patch characteristics changed significantly between periods. Within the time span analyzed for the study area, we conclude that increased fire occurrence promoted landscape heterogeneity while fire exclusion did not. Published by Elsevier B.V.

show abstract

“…Principal component analysis (PCA) is a multivariate statistical technique that selects uncorrelated linear combinations (eigenvector loadings) of variables in such a way that each successively extracted linear combination or principal component (PC) has a smaller variance 21 . The eigenvector matrix used to perform PCA for each subset was examined to identify which PC contained the target (mineral) information.…”

Section: Principal Component Analysis Methodsmentioning

confidence: 99%

Untitled

Myint¹,

Won-In²,

Takashima³

et al. 2008

ScienceAsia

View full text Add to dashboard Cite

Landsat 7 Enhanced Thematic Mapper Plus image data was used to identify and map lateritic soil zones in the Phrae basin which is one of the largest intermountain basins in northern Thailand. The lateritic soil zones were discriminated using band ratio and principal component analysis. The lateritic soil detection images were processed by band ratio (band 3 / band 1), principal component analysis of bands 1 and 3, and principal component analysis of bands 1, 3, 4, and 5. The results of these three indices were superimposed using GIS to define a preliminary lateritic soil image of the study area. A threshold method was used for converting a grey scale image into a binary image. Different threshold values were used to find the most probable areas of lateritic soil zones in the image. The threshold values were determined from a published geological map and known lateritic soil areas with good exposure in the image. The quality of the results was evaluated by the normalized difference vegetation index. Field investigation was carried out to substantiate the remote sensing investigation and the laboratory GIS analysis. This method can also be applied to other lateritic soil and iron oxide regions.

show abstract

Standardized principal components

Cited by 325 publications

References 8 publications

Detection of fire impact and vegetation recovery over tropical peat swamp forest by satellite data and ground-based NDVI instrument

Detection of fire impact and vegetation recovery over tropical peat swamp forest by satellite data and ground-based NDVI instrument

Trends in fire patterns in a southern African savanna under alternative land use practices

Untitled

Contact Info

Product

Resources

About