Advanced search
Start date
Betweenand

Generation of DSM based on segmentation and classification of hyperspectral image

Grant number: 14/24844-6
Support Opportunities:Scholarships abroad - Research Internship - Doctorate
Effective date (Start): March 05, 2015
Effective date (End): March 04, 2016
Field of knowledge:Physical Sciences and Mathematics - Geosciences - Geodesy
Principal Investigator:Antonio Maria Garcia Tommaselli
Grantee:Raquel Alves de Oliveira
Supervisor: Eija Honkavaara
Host Institution: Faculdade de Ciências e Tecnologia (FCT). Universidade Estadual Paulista (UNESP). Campus de Presidente Prudente. Presidente Prudente , SP, Brazil
Research place: Finnish Geospatial Research Institute (FGI), Finland  
Associated to the scholarship:13/17787-3 - DEVELOPMENT AND ASSESSMENT OF TECHNIQUES FOR GENERATION OF DENSE POINTS CLOUD USING HYPERSPECTRAL IMAGES ACQUIRED BY UAV, BP.DR

Abstract

Digital Surface Models (DSM) of forest areas can be derived from digital images and the resulting accuracy and completeness will depend on the image resolution and quality. Image matching is the main step in this process of DSM generation but, despite having been extensively explored in different areas, there is a room to improve these techniques in forest scenarios. Hyperspectral images give detailed spectral information for each pixel in an image, which aids the segmentation and classification process. Classified images can improve the process of image matching, driving the tuning of the parameters of matching strategies for different objects (classes) and providing key points for feature based matching. Most of the matching algorithms uses single or tri-band (RGB) images and the extension to multispectral or hyperspectral images is a more complex process due to the differences between spectral responses of bands and the amount of data to be processed. In this project, unlike classic hyperspectral sensors based on pushbroom scanning, hyperspectral images will be acquired in frame format, favorable to unstable platforms. Nevertheless, the frame camera to be used does not acquire all bands instantaneously, causing band misregistration due to the platform motion. In this context, the main aims of this project with the Finnish Geodetic Institute (FGI) are: to obtain experience with the Remote Sensing methods developed by the FGI group for Finnish forests; to study and develop techniques for hyperspectral image registration; to evaluate the appropriate segmentation and classification methods for high-resolution images in remaining Atlantic forests; and to evaluate the proposed method using different data sets. Quality assessment will be performed comparing DSM generated with different software used in the FGI and the software under development in this project. (AU)

News published in Agência FAPESP Newsletter about the scholarship:
Articles published in other media outlets (0 total):
More itemsLess items
VEICULO: TITULO (DATA)
VEICULO: TITULO (DATA)

Scientific publications
(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)
DE OLIVEIRA, RAQUEL A.; TOMMASELLI, ANTONIO M. G.; HONKAVAARA, EIJA. Geometric Calibration of a Hyperspectral Frame Camera. PHOTOGRAMMETRIC RECORD, v. 31, n. 155, p. 325-347, . (13/50426-4, 14/24844-6, 13/17787-3)
OLIVEIRA, RAQUEL A.; TOMMASELLI, ANTONIO M. G.; HONKAVAARA, EIJA. Generating a hyperspectral digital surface model using a hyperspectral 2D frame camera. ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, v. 147, p. 345-360, . (14/24844-6, 13/50426-4, 13/17787-3)

Please report errors in scientific publications list by writing to: cdi@fapesp.br.