International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B3, 2012 | tornato
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia |
| Clausi, D
Identified ground truth features | Gompanisor
Classification Method % Area occupied Markov Ra
by the class as | Confence
per classification Agriculture Built-up area | Water bodies http;//ieeext
results (100%) (100%) (100%) Chellappa,
Lucieer et al’s Agriculture 48.48 78.26 41.64 m M
LBP analysis and 959-963
ISODATA m
(Figure 2b) Built-up area 36.83 12.23 -
Chen, C H.
Pattern F
Water bodies 14.69 9.5] 58.36 qi Scie!
Proposed LBP Agriculture 82.55 19.16 13.8 Dog 7.
analysis and Js SSH
CD Built-up area 13.66 73.65 - 329.
Water bodies 3.79 7.19 86.20 m à
Haralick, |
Textural fi
Table 1: The comparative success rate for classifying the features obtained by applying “Lucieer et on systems,
al’s LBP analysis and ISODATA" and “Proposed LBP analysis and ISODATA” separately on
RISAT-II X-Band image. The column of the Table represents the percentage (%) of area occupied Jain, A. K.,
by the features according to the classification results when there is a unique feature as per the ground review, AC
truth.
Kohei, A.,
with param
- The experimental results with the input image (Figure 2a) A
The experimented results with the input image (Figure 2a) shows that the use of “Lucieer et al's" technique can http;/dlww
shows that the “Lucieer et al’s LBP analysis and ISODATA” superpose regions namely built-up area and agriculture as wac jp/con
technique under segment (i) agriculture area mixed with built- shown in Figure 2b. The proposed technique mostly January 17,
up area, (ii) water bodies mixed with the agriculture shown in overcomes these discrepancies as shown in Figure 2c.
Figure 2b. This discrepancy decreases the success rate of Lucieer, A
recognizing agriculture, built-up area and water bodies as segmentati
shown in Table 1. The “Proposed LBP analysis and 3.0 CONCLUSIONS identificatic
ISODATA” mostly overcome these discrepancies. It is found Sensing, 26
that the superposition of agriculture, water bodies, and built-up In this paper LAM (Local Adaptive Median) filter is developed
area becomes less as shown in Figure 2c. Moreover the to suppress the speckle noise from RISAT-II image. LBP is Oliver. C.
decreased discrepancies increase the success rate in recognizing used as a tool to compute the degree of texture around each Ape The R
agriculture, water bodies and built-up area (shown in Table 1). pixel in the microwave image. This computed texture measure
around each pixel in the image is used farther to classify the Ojala, T.P.
microwave image. From the results of the experiments it is i a
2.4.1 Comparison between Lucieer et al's [2005] found that the proposed method adequately clusters complex distribution
classification technique and proposed technique images containing texture region as well as non-texture region.
Moreover it can be considered as an intuitively appealing, Ojala, T
unsupervised (no need for a predefinition of the number of Multiresol
- Lucieer et al’s employ a circle of fixed radius ‘c’ from the clusters) and fast clustering algorithm. As a result the method is classificati
center pixel position of the kernel and. intersected pixels on the potentially useful to classify RISAT-II microwave images Pattern An
perimeter of that circle are only considered for measuring the efficiently and accurately.
texture around that pixel. As a result most of the pixels of the Petrou, M.
kernel are not used for measuring the texture. The proposed histograms
method uses a series of circles (2D) centered on the pixel with 4.0 REFERENCES Lett., 23(9)
incremental radius values. The intersected pixels on the
perimeter of the circles of radius r (where r = 1, 3 and 5) are Bernad, P., Denise, G., and Réfrégier, R., 2009. Hierarchical Richards
considered for measuring the texture around that pixel. As a feature-based classification approach for fast and user- Analysis: A
result it uses more number of pixels of the kernel for measuring interactive SAR image interpretation, IEEE Geosci. Remote
the texture around each pixel of the image. Sensing Letter. 6(1), pp. 117-121. Szira yi, T.
2000.
- Lucieer et al's used supervised (fuzzy c-means) classification Bezdek J C., Ehrlich R and Full W., 1984. FCM: the Fuzzy c- in C m.
technique where as the proposed technique an un-supervised Means clustering algorithm, Computers and Geosciences, 10, Imaging, 6
(ISODATA) classification technique. pp. 191-203. 3
