Object-Oriented Image Segmentation and Classification

IDRISI includes a suite of tools to support object-oriented segment-based classification. Segmentation is a process by which pixels in one or more images are grouped into segments, or objects, that share a homogenous spectral similarity. Object-oriented segment-based classification is an approach that classifies remotely-sensed images and ancillary data based on these objects rather than individual pixels.

Objected-oriented segment-based classification is highly suited for applications that utilize medium to high resolution satellite imagery and is a useful addition for those mapping land cover and monitoring land change. Other applications, such as biodiversity and habitat mapping, can also take advantage of this classification approach.

The object-oriented segment-based classification approach within IDRISI is straightforward and bypasses the complex rule-base construction process found in other tools. Once identified, the objects can be used as input to the powerful set of existing classification routines in IDRISI, in particular the wide range of machine learning tools such as neural network classifiers or classification tree analysis.

Learn more about IDRISI Selva

Segmentation Classification Focus Paper (PDF)
IDRISI Selva Brochure (PDF)
Image Processing in IDRISI

 
Image Segmentation and Forest Crown Segments using IDRISI

The  SEGMENTATION module in IDRISI creates an image of segments that have spectral similarity across many input bands. The image on the left is a false color composite image of a heavily forested area with high species diversity. The image on the right is a result of segmentation on the original multispectral bands that produce crown-level segments.

Back to Top

 
Segmentation Classification IDRISI-s

The module SEGCLASS classifies the imagery using a majority rule algorithm to assign each segment to the majority class from the reference image. SEGCLASS can improve the accuracy of a pixel-based classification and produce a smoother map-like classification result while preserving the boundaries between the segments.

Back to Top

Clark University