Extraction of Vegetation Image Index Using Normalized Difference Vegetation Index Algorithm
Abstract— The study aims to create an application using openCV that can identify and calculate the vegetation part of a satellite image. The study utilized the two out of seven Landsat-8 band images over a part of Metro Manila in Philippines acquired on Feb. 13, 2016. These band images are Bands 4 and 5 which are normally used to compute the vegetation index of a satellite image. The algorithm calculates the relative area of the vegetation using Normalized Difference Vegetation Index or NDVI formula. The output of the NDVI creates a single-band dataset that only shows greenery. Values close to zero represent rock and bare soil and negative values represent water, snow and clouds. Taking ratio or difference of two bands makes the vegetation growth signal differentiated from the background signal. Water has an NDVI value less than 0, bare soils between 0 and 0.1, and vegetation over 0.1. Increase in the positive NDVI value means greener the vegetation. The sample satellite image is Manila with Landsat-8 operational land imager (OLI) images. The study aims to test the NDVI formula in extracting vegetation index of Manila region which can be used to monitor the urban and classification of a certain region.
Index Terms— remotes sensing, ndvi, satellite image
Dr. Ace C. Lagman, Joferson L. Bombasi
FEU Institute of Technology, PHILIPPINES
Far East University Korea, KOREA
Cite: Dr. Ace C. Lagman, Joferson L. Bombasi, Chul-Soo Ye, "Extraction of Vegetation Image Index Using Normalized Difference Vegetation Index Algorithm," Proceedings of 2018 the 8th International Workshop on Computer Science and Engineering, pp. 730-734, Bangkok, 28-30 June, 2018.