eCognition: Object Based Image Analysis (OBIA) software available in the GIS-studio

            eCognition Developer is a powerful development environment for object-based image analysis. It is used in earth sciences to develop rule sets for the automatic analysis of remote sensing and other data. eCognition Developer can be applied for all common remote sensing tasks such as vegetation mapping, feature extraction, change detection and object recognition. The object-based approach facilitates analysis of all common data sources, such as medium to high resolution satellite data, high to very high resolution aerial photography, lidar, radar and even hyperspectral data.
           

How eCognition Developer works

           In the data loading step, data can be imported in arbitrary combinations. Different resolutions are supported as well as differing sensor types. Image segmentation derives the homogeneous image regions that provide the base layer for the following analysis steps. Analysis can be implemented based on conditions, samples or a combination of both. Through context analysis, features that are not identifiable solely based on spectral or textural attributes can be extracted. The extracted features can be exported in raster or vector format allowing smooth integration into GIS workflows. Rule sets and applications developed for one task can be reused over large areas, effectively automating image analysis.

eCognition used in a course module

           Traditional software uses spectral signatures for pixel-based classifications of a satellite image into predefined categories with homogeneous spectral characteristics. Basically such pixel- based classification can be separated into unsupervised and supervised classifications. The eCognition software in contrast, is based on object-oriented classification. This means that an raster image is first segmented into coherent / homogeneous objects. The size of these objects, also called segments, can be manipulated by the interpreter using parameters such as shape and compactness and color. The actual classification takes place after the segmentation into objects. The eCognition software makes it possible to rapidly subdivide a satellite image or an air photo into homogeneous units of recognizable shape, color, and texture. It is ideal for analyzing human influenced and / or highly fragmented landscapes. http://www.gis-studio.nl/index.php?page=msc#jesus

 

eCognition used in a publication

Name: Jesus Aguirre, Harry Seijmonsbergen and Joost Duivenvoorden M.Sc Thesis Year: 2010	Optimizing land cover change detection using combined pixel-based and object-based image classification in a mountainous area in Mexico A combination of object-based and pixel-based image classification methods and post-classification object-based change detection is applied to Enhanced Landsat Thematic Mapper images for optimizing land cover classification and change detection in a mountainous region in northern Mexico. The land cover categories with the highest individual classification accuracies for coniferous forest, shrubs and deciduous forest, bare soil, urban areas and water bodies in the object-based and pixel-based classifications were extracted and combined into single combined classification layers. Comparison of the overall classification accuracies of the object-based, pixel-based and the combined maps shows that the combination method produces higher overall accuracies. The combined classification maps are used as input for post-classification object-based change detection analysis. The combination of pixel-based and object-based change detection method for Landsat imagery leads to improved classification and change detection classification and has potential to similar mountain areas. Figure shows examples of classifications of a Landsat satellite image of 1999 and 2006 and the changes detected using a post-classification technique. Please click here for a full-sized figure. Back to top
	admin | log out