We propose a method of segmentation of remote sensing time series data, which exploits multi-temporal information to identify objects’ boundaries. Extracting homogeneous objects with similar temporal behavior, the method analyzes large volumes of multi-temporal input data in a piecewise way and produces a consistent output segmentation layer for large territories. Segment building logic is simplified to minimize the computation time, while objects’ boundary identification accuracy remains sufficient for remote monitoring and mapping of vegetation, and specifically, agricultural crops. At the Space Research Institute of the RAS, the proposed method is currently applied for automated on-line satellite imagery analysis for recognition and mapping of (winter and spring) crops on large territories and land-use evaluation. The method successfully deals with gaps in remote sensing time series data and performs well even when input images are contaminated with speckle noise. Due to its ability to map dynamically homogeneous surface areas with partially missing data, the method provides a potential for their recovery.
Работа посвящена развитию методов картографирования растительного покрова России на основе данных дистанционного зондирования Земли из космоса. Представлена первая версия карты растительного покрова Российской Федерации на основе данных спутниковой системы Proba-V. В качестве признаков распознавания типов земного покрова при картографировании использовались очищенные от влияния облаков и некоторых других мешающих факторов сезонные композитные изображения, полученные по данным Proba-V 2016 г. с пространственным разрешением 100 м, характеризующие, прежде всего, фенологическую динамику спектрально-отражательных характеристик различных классов растительности. Для распознавания типов земного покрова по их спектрально-отражательным характеристикам использовался метод локально-адаптивной классификации LAGMA и соответствующий программный комплекс обработки данных дистанционного зондирования LAGMA-PLUS. Локально-адаптивная классификация выполнялась с применением алгоритма максимального правдоподобия. Легенда полученной карты включает в себя 23 класса земного покрова, образующих шесть различных тематических групп. Использование данных системы Proba-V демонстрирует значительное повышение детальности картографирования растительного покрова по отношению к аналогичным картам, созданным ранее по данным системы MODIS пространственного разрешения 250 м. Наличие постоянно обновляемого открытого архива многолетних данных Proba-V открывает возможность ежегодного картографирования растительного покрова России.
In this study we present methods and automatic technology developed for routine processing of satellite imagery acquired by cameras MSU-201 and MSU-202 (KMSS-M) onboard Meteor-M №2. The developed methods were aimed at imagery georeferencing issues fixing, clouds and shadows detection as well as atmospheric and radiometric correction. Basing on these methods we built an automatic technology and complete KMSS-M data processing chain which provided analysis ready dataset for Russian grain belt and adjacent areas of neighboring countries for the year 2020. Method for imagery georeferencing was based on Pearson’s correlation localized maximization when compared to the georefenced and cloudfree coarse-resolution reference image produced in IKI RAS through MOD09 product time series processing. Method for clouds and shadows detection was based both on the spatial analysis of outputs from geocorrection step and auxiliary image, characterizing georeferenced KMSS-M image values relative accordance with the IKI reference image. The atmospheric correction was based on localized histogram matching of KMSS-M and IKI reference date-corresponding imagery, and thereby concurrently performed radiometric correction of KMSS-M data, compensating effects of varying viewing and illumination geometry which explicitly manifest across 960-km-wide swath area. The developed methods are noticeably minimalistic, requiring only one target spectral band to perform properly. Due to high flexibility and robustness, they also may be applied to raw satellite imagery acquired from various Earth observation systems, including Russian systems of high and moderate spatial resolution. The technology is currently being deployed in an operative mode for several test sites of Russia since the year 2021 onwards.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.