Page 52 of
| Previous | Next |

| TOC |

Results and Disscussions

1.3.4. Quantitative Evaluation of Supervised Classification of the Images

Accuracy assessment determines the quality of the information derived from the remotely sensed data. Through quantitative accuracy assessment we can identify and measure map errors. Accuracy assessment was done for the LANDSAT, LISS III and ASTER classified images. Quantitative accuracy assessment involves the comparison of a site on the map against reference information for the same site. Accuracy requires:

1. Design of unbiased and consistent sampling procedures.
2. Rigorous analysis of the sample.

The producer’s accuracy denotes how well a particular class can be classified in an image. Users accuracy denotes that the pixels present in a particular class actually belongs to an another class.

Accuracy Assessment is carried out for all the three data sets. The error matrix tables were shown in Annexure 4.

The Table 6. and 7 shows producers, users and overall accuracy.

Table 6: Accuracy Assessment



Table 7: Table showing the Accuracy Results of the LANDSAT, LISS III and ASTER Images

| Previous | Next |