ABSTRACT:The task of crop type classification with multitemporal imagery is nowadays often done applying classifiers that are originally developed for single images like support vector machines (SVM). These approaches do not model temporal dependencies in an explicit way. Existing approaches that make use of temporal dependencies are in most cases quite simple and based on rules. Approaches that integrate temporal dependencies to statistical models are very rare and at an early stage of development. Here our approach CRF multi , based on conditional random fields (CRF), should make a contribution. Conditional random fields consider context knowledge among neighboring primitives in the same way as Markov random fields (MRF) do. Furthermore conditional random fields handle the feature vectors of the neighboring primitives and not only the class labels. Additional to taking spatial context into account, we present an approach for multitemporal data processing where a temporal association potential has been integrated to the common CRF approach to model temporal dependencies. The classification works on pixel-level using spectral image features, whereas all available single images are taken separately. For our experiments a high resolution RapidEye satellite data set of 2010 consisting of 4 images made during the whole vegetation period from April to October is taken. Six crop type categories are distinguished, namely grassland, corn, winter crop, rapeseed, root crops and other crops. To evaluate the potential of the new conditional random field approach the classification result is compared to a manual reference on pixel-and on object-level. Additional a SVM approach is applied under the same conditions and should serve as a benchmark.
ABSTRACT:In many publications the performance of different classification algorithms regarding to agricultural classes is evaluated. In contrast, this paper focuses on the potential of different imagery for the classification of the two most frequent classes: cropland and grassland. For our experiments three categories of imagery, high resolution aerial images, high resolution RapidEye satellite images and medium resolution Disaster Monitoring Constellation (DMC) satellite images are examined. An object-based image classification, as one of the most reliable methods for the automatic updating and evaluation of landuse geospatial databases, is chosen. The object boundaries are taken from a GIS database, each object is described by means of a set of image based features. Spectral, textural and structural (semivariogram derived) features are extracted from images of different dates and sensors. During classification a supervised decision trees generating algorithm is applied. To evaluate the potential of the different images, all possible combinations of the available image data are tested during classification. The results show that the best performance of landuse classification is based on RapidEye data (overall accuracy of 90%), obtaining slightly accuracy increases when this imagery is combined with additional image data (overall accuracy of 92%).
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.