Computer-Aided Intelligent Recognition

The software adopts a expert decision-support system and a fuzzy classification algorithm to extract information based on the context of a semantic network hierarchy, breaking with the traditional commercial software based solely on spectral information from pixels for image segmentation limitation. It uses the most advanced classification technique - Object-Oriented classification method, for improved accuracy in automatic recognition of features from remote sensing data at different spatial resolutions, effectively meeting user needs.

Multi-source data integration

The software integrates earth observation image data such as the World view, Landsat, Spot, QuickBird, ZY-3, SAR, aerial imagery, LIDAR and GIS data at different spatial-temporal resolutions,. Simultaneous integration of multi-source data during remote sensing information extraction ensures accuracy.

Efficient multi-scale image segmentation

Any type of panchromatic or multi-spectral data at a selected scale is divided into homogeneous image objects, creating an image object hierarchy network. In the object hierarchy, a small object is a child of large objects. Each object has its own context, neighbors, child and parent objects. Image objects and object attributes can be used for further classification based on the relationships between objects.

Sample-based supervised classification

GEOWAY Feature offers a simple and fast classification function by selecting, and training samples for supervised classification. The classification algorithms include K-nearest neighbor, support vector machine, Bayesian classification, and decision tree classification algorithms.

Knowledge-based fuzzy logic classification

GEOWAY Feature uses an inheritance mechanism, fuzzy logic, and a semantic model for classification of knowledge; for example, by defining a NDVI, DNWI knowledge base to extract water and vegetation from imagery.

Vector-raster-based integrated interacting editing

GEOWAY Feature gives full play to the advantages of integrated vector and raster editing to meet production and engineering applications needs with efficient and accurate remote sensing information extraction.

This software is used for the investigation of natural resources and the environment, agriculture, forestry, land use, pipeline management, telecommunications, urban planning, cartography, natural disaster monitoring, coastal and ocean mapping, and mining applications.