Remote Sensor Object Based Image Analysis or ‘RSOBIA’ is a NEW extension to Maritime Toolbar for Esri ArcMap, which takes raster imagery and segments the data into geographic areas with similar statistical properties.
It interprets and digitises the areas and boundaries seen in imagery (such as sidescan sonar and multi-beam backscatter) and creates a polygonised map of the results (e.g. areas of similar seafloor type). It supports Segmentation and Classification which are key techniques for image analysis, and gives quick, easy, repeatable results.
Without the tool, digitising images can be slow, subjective – dependent on the interpreter – and is non-repeatable. Using RSOBIA, segmentation of the results will provide straight-forward and objective polygons which are statistically calculated. Imagery derivative processing techniques are provided for ease of use but also include texture analysis techniques such as homogeneity, dissimilarity, contrast and others (Grey Level Co-occurrence Matrices – GLCMs). Speed of results will depend on the size of the image and speed of the computer being used.
Delivered as a part of Maritime Toolbar, RSOBIA is compatible with ESRI ArcMap 10.0 to 10.4 making it integral to your workflow.
This extension has been created by OceanWise’ Licensed Partner Tim Le Bas and is maintained and supported by OceanWise as a collaborative project.
Tim Le Bas is a Senior Marine Habitat Scientist for the NOC and has been working in the marine environment, mapping the seafloor for over 30 years. The emergence of available GIS systems has allowed numerical and statistical interpretation techniques to be developed. Tim developed the RSOBIA software to work with GIS systems and reduce the time spent laboriously classifying marine images by hand. This process which could take hours, days or even weeks by hand, can now be done in minutes using RSOBIA.