BRIN provides ARD data to fulfil user request so that users can use the data directly. This greatly helps users in analyzing remote sensing data without preprocessing. One of the ARDs produced by BRIN is a tile-based mosaic for the Pleiades - very high-resolution satellite imagery. We proposed the color balancing method to balance the color and brightness of the mosaic images. The method of histogram matching for the Pleiades image is developed based on the Sentinel-2 image as the reference image in the process. We assessed the effect of color balancing on land cover classification accuracy to show the reliability of the color balancing method. A comparison of the land cover classification before and after color balancing was conducted in the experiments. As a result, by using the Support Vector Machine, the classification result after color balancing has higher accuracy than before. The overall accuracy is 91% and it is 50% better than the classification results for the classification result without the color-balancing process. Thus, the proposed method of color balancing is an alternative method to balance the color and brightness of the mosaic images.
The effect of color balancing on the accuracy of land cover classification for pleiades mosaic data
THE 9TH INTERNATIONAL SEMINAR ON AEROSPACE SCIENCE AND TECHNOLOGY – ISAST 2022 ; 2022 ; Bogor, Indonesia
AIP Conference Proceedings ; 2941 , 1
2023-12-11
9 pages
Conference paper
Electronic Resource
English
Data mining from PLEIADES telecommand logbooks
AIAA | 2014
|Online Contents | 2012
Online Contents | 1998
Online Contents | 2012