Geological Applications of LANDSAT Thematic Mapper Imagery:
Mapping and Analysis of Lineaments in NW Penisula Malaysia
Data and Study area
Landsat TM data which are recorded in 1989 covering part of frame 128/56 in the international references system are used in this study. That data were recorded in computer compatible tape-CCT, covering part of Kedah state of approximately 2,500 km square (Figure 1). The geological map of the area (Figure 2) showing the most of the area is covered by metamorphic and intrusive igneous rocks. The metamorphic rock is known as Baling Formation of Ordovician-Silurian, while the intrusive igneous is Early-Middle Cretaceous to Early Jurassic age (Hutchison, 1973). Apart from that, the area are also consists of sedimentary rocks of Triassic Semanggol Formation which are effected by low grade metamorphism and Jurassic-Cretaceous rock formation (Burton, 1973). Major fractures and faults are the main structural features shown on the geological map of the area.

Figure 1: Location map of the study in the northwest Peninsula Malaysia

Figure 2: Slightly simplified map of solid rock geology in the study area (after Geological Survey
of Malaysia. 1985).
Data processing and image products
Following a number of processing techniques (Jensen, 1986 and Mather, 1987), the contrast of six bands data were digitally enhanced. The images of all bands were compared in terms of contrast and definition of geological features (lineaments). As a result of visual evaluation, Landsat TM band-4 data which record the information at the wave-length between 0.75 to 0.90 mm were selected for this study, since it shows good contrast and display geological lineaments compared to the other bands.
In order to enhance the structural geological information (lineament) further, filtering techniques were applied to this band-4 Landsat TM data. The result shows that the a number of filters of the following values produced very good images enhancing the lineaments of the area. The filter values are as follows.
- Non-directional filter (Laplacian convolution filter)
0 -1 0
1 5 -1
0 -1 0
- Directional filter
| 1 |
0 |
1 |
|
-1 |
-2 |
-1 |
| 2 |
0 |
2 |
0 |
0 |
0 |
| 1 |
0 |
1 |
1 |
2 |
1 |
| (north-south) |
(east-west) |
| |
| 2 |
-1 |
0 |
0 |
-1 |
-2 |
| 1 |
0 |
1 |
1 |
0 |
-1 |
| 0 |
1 |
2 |
2 |
1 |
0 |
| (northeast-southwest) |
(northwest-southeast) |
Therefore, the following images are used in this study, that is:
- Image of Band-4 which contrast is digitally enhanced
- Image of band-4 which is filtered by Laplacian convolution filter.
- Four images of band-4, each one is filtered by directional filer N-S, E-W, NE-SW and NW-SE respectively.