Landcover Mapping using Landsat Image

Agus ArisG051110011

INFORMATION TECHNOLOGY FOR NATURAL RESOURCE MANAGEMENT

SEAMEO – BIOTROPBOGOR AGRICULTURAL UNIVERSITY

2011

Introduction Remote sensing is the science and art of obtaining information about an

object, area, or phenomenon through the analysis of data acquired by a device that is not in contact with the object, area, or phenomenon under investigation

Remote sensing is one method of observation or measurement of the spatial elements of the earth's surface. method is very effective and efficient has many variants in providing spatial data records, then to improve quality and enhance sensors recording data related to the production end of the spatial data (hardcopy), carried out the processes of digital image processing. In this connection we also use the device ErDas or Er Mapper

Erdas is one of the benefits for image processing in the process of classifying the coastal area or an area of land cover

Objective

Make a land cover using Software Remote sensing

Measure land cover area in Central Java

Data Requirement

Sensor Name Spatial Resolusi

(m)

SpectralResolusiNumber of Band & wavelength (µm)

Radiometrik Resolution (bit)

Temporal Resolution

(day)

launch time

LANDSAT 7 ETMMultispectral-Reflective-ThermalPan

30 m90 m15 m

Band 1-5&7 : 0.4 – 2.35Band 6 : 10.4– 12.5Band 8 : 0.52 – 0.90

888

16 days 1998

Image Processing Composit Georeference ( Geometric

Correction ) Masking Ground check using Googel Earth Make AOI (Area Of Interest) Classification Supervised

Composit

Before After

Composit is how do we combine all the bands

Georeference

Before After

Georeference is how to make same the coordinate of the image with the actual conditions on the ground

MaskingMasking is a step of being done to eliminate some areas that are not needed

Ground Check

Criteria of visual interpretation

Color Tone Size Pattern Texture Shape Shadow Association

AOI Porpose AOI using to make the area

for classification

Accuracy assesment

2000 Exelent1900 - < 2000 Good1800 - < 1900 Fair1600 - < 1800 Poor< 1600 Inseparable

overall accuracy 88.23895457

kappa accuracy 81.12461401

Evaluate Contingency Evaluate Separability

Classification SupervisedA method used to group classes defined by the user by using the Training Area

Area

Area = Number of Pixel / Resolusi