Advanced search
Start date
Betweenand

Macrophyte analysis and classification using computer vision techniques and bio-optical models from integration of satellite and drone multispectral images

Grant number: 19/16698-3
Support type:Research Grants - Innovative Research in Small Business - PIPE
Duration: July 01, 2020 - March 31, 2021
Field of knowledge:Physical Sciences and Mathematics - Geosciences
Principal Investigator:Alisson Fernando Coelho do Carmo
Grantee:Alisson Fernando Coelho do Carmo
Empresas:Empresa a definir
Inspectral Soluções Inovadoras em Tecnologia da Informação Espacial
CNAE: Consultoria em tecnologia da informação
Pesquisa e desenvolvimento experimental em ciências físicas e naturais
City: Presidente Prudente
Co-Principal Investigators:Nariane Marselhe Ribeiro Bernardo do Carmo
Assoc. researchers:Ana Carolina Campos Gomes ; Enner Herenio de Alcântara ; Fernanda Sayuri Yoshino Watanabe ; Milton Hirokazu Shimabukuro ; Nilton Nobuhiro Imai
Associated scholarship(s):20/10293-9 - Developing methods to identify macrophytes in hydroelectric reservoirs using computer vision techniques and spectral indexes using satellite and drone images, BP.TT
20/10315-2 - Using computer vision and machine learning algorithms with satellite images integrating with drone images for macrophyte detection, BP.TT
20/08484-0 - Analysis and classification of macrophytes using computer vision techniques and bio-optical models applying in drones and satellite images, BP.PIPE

Abstract

Lakes and rivers are vital ecosystems that play fundamental role in water supplies for human and animal consumes, agriculture production, industrial activities and hydroelectric generation, besides, water are also essential to recreation, navigation and fishery activities. Overall, mainly in urban and well developed areas, it is needed to preserve and manage our aquatic systems. However, the actions that aim to reduce the wasting water are not enough when compared to the population increase rates and the increase of agricultural and industrial activities, that are directly related to the pollution sources of the eventual pollution sources of aquatic resources. The run off of herbicides from agriculture and the huge discharges of wastewater without treatment are the main causing of disturbance in environmental equilibrium, increasing the availability of nutrients that implies to algae blooms and accelerated growth of aquatic plants to undesirable levels. At certain levels, native aquatic plants are relevant because they represent an oxygen and food source, as well as, can developed a habitat function to some aquatic organisms. When we have an accelerated growth rate, the presence of aquatic plants can cause difficulties in the mobility of navigation, fishery, recreation and even reduce the potential of hydroelectric production. Regarding to all these consequences, the most evident problems related to aquatic plants are the issues related to navigation and hydroelectric generation. It is already related that some hydroelectric and waterways degraded their performances due to aquatic plants infestation by emerged, submerged and/or floating plants. In front of that, traditional techniques as in situ monitoring cannot be conducted because the plants occupied, some times, the entire reservoir and did not allow the boats displacements. In this context, remote sensing techniques offers an alternative way to characterize and monitor aquatic systems, using in situ observations as reference dataset that can validate the experiments. Therefore, remote sensing represents a relevant tool for environmental monitoring due to temporal (high frequency to produce information) and spatial (broader areas covered by an image) resolutions, reducing the costs of field measurements. Dynamic ecosystems as hydroelectric reservoirs highlighted the frequent monitoring against the fast and often changes of catchments and their impacts into the water conditions. In this context, this project investigates models and techniques that allows to monitor the macrophyte dynamics in aquatic systems, mainly in rivers and tributaries that are part of reservoirs, using remote sensed dataset captured by orbital and aerial sensors (Drones). The future solution, relied on the evidence of the technical-scientific viability resulted from the present project, will allow creating systematic mechanisms to monitor macrophyte using remote sensed images, as well as, reevaluate the mobility of macrophytes between reservoirs and identify what are the effects when a sustainable management is provided. The project also corroborates to create automatic, quick and precise reports according to the user's demands and manages all steps of image processing considering all remainder variables evolved in this process. (AU)