Occlusion Detection by Height-Gradient-Based Method: Refi nement of Result Using Mathematical Morphology

Main Article Content

Henrique Cândido de Oliveira
Aluir Porfírio Dal Poz
Mauricio Galo

Abstract

Nowadays, the use of orthophotos in projects related with infrastructure and urban plan is frequent. Urban areas usually
have many tall buildings that cause occlusions in aerial images. These occlusions can be severe depending on the
height of the buildings, image acquisition viewpoint, and due to the use of digital terrain models, which represent only
the ground in the orthophoto generation. The occlusion detection method applied in this project uses a digital surface model for height gradient calculation, instead of using a digital building model as an initial information. Analysis of
the surface height-gradient at certain sampled directions allows the identifi cation of occluded regions. The aim of this
paper is to assess the use of mathematical morphology to refi ne results of occlusion detection in aerial images for true
orthophoto generation. In the assessment, the dilation and closing operators were compared. For the experiment, a
data set composed of aerial images and point cloud from Presidente Prudente-SP / Brazil, with density around 8 points/
m2, was used. The result was analyzed by quantitative and qualitative indexes, i.e., the completeness and the correctness.
Regarding the correctness, both operators presented similarity with a slight advantage for closing operator. On the other
hand, the experiment performed with dilation operator results in a completeness index of approximately 6% greater
when compared with closing operator. This advantage indicates that dilation is a suitable operator for the purpose of
occlusion detection refinement.

Downloads

Download data is not yet available.

Metrics

Metrics Loading ...

Article Details

How to Cite
DE OLIVEIRA, H. C.; DAL POZ, A. P.; GALO, M. Occlusion Detection by Height-Gradient-Based Method: Refi nement of Result Using Mathematical Morphology. Revista Brasileira de Cartografia, [S. l.], v. 67, p. 1663–1674, 2019. DOI: 10.14393/rbcv67n0-49240. Disponível em: https://seer.ufu.br/index.php/revistabrasileiracartografia/article/view/49240. Acesso em: 22 jul. 2024.
Section
Artigos

Most read articles by the same author(s)

1 2 > >>