Application of different filters for smoothing and noise removal of satellite images time series

• Grant number: 18/13295-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: August 01, 2018
• End date: April 30, 2020
• Field of knowledge: Agronomical Sciences - Agricultural Engineering
• Principal Investigator: Rubens Augusto Camargo Lamparelli
• Grantee: Ana Flávia Carrara Bonamigo
• Host Institution: Faculdade de Engenharia Agrícola (FEAGRI). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated research grant: 14/26767-9 - Global Sustainable Bioenergy Initiative: geospatial & environmental analysis of pastureland intensification for bioenergy, AP.SPEC

Abstract

There are still few studies on the dynamics of crop-livestock integration systems. Geographical distribution is one of the major challenges for the monitoring of these mixed systems due to their characteristics, since agricultural crops are planted in rotation or in a consortium, making it difficult to identify these targets by means of satellite images. The use of multitemporal series of vegetation index of a cultivated area allows to follow the development of a given crop and also to compare different behaviors over the years. These multitemporal series can provide key elements in the spectral discrimination of different crops in the field. To eliminate abrupt variations in the vegetation index values of the multitemporal series commonly caused by cloud presence, angle of view, among other factors, it is necessary to filter the data in order to soften the data and obtain better results. Based on this, the objective of this work is to test different filters in a multitemporal series of vegetation index and to verify which filter best fits the data and the integration system.