Mseg: a Generic Region-based Multi-scale Image Segmentation Algorithm for Remote Sensing Imagery

The objective of this research was the development of a generic image segmentation algorithm, as a low level processing part of an integrated object-oriented image analysis system. The implemented algorithm is called Mseg and can be described as a region merging procedure. The first primitive object representation is the single image pixel. Through iterative pairwise object fusions, which are made at several iterations, called passes, the final segmentation is achieved. The criterion for object merging is a homogeneity cost measure, defined as object heterogeneity, and computed based on spectral and shape features (indices) for each possible object merge. The heterogeneity is then compared to a user defined threshold, called scale parameter, in order for the decision of the merge to be determined. The processing order of the primitive objects is defined through a procedure (Starting Point Estimation), which is based on image partitions, statistical indices and dithering algorithms. Mseg provides several parameters to be defined by the end user. Mseg offers a multi-resolution algorithm which performs segmentations at several levels, and at the same time provides automatic topology of objects within each level and among levels. The algorithm was implemented in C++ and was tested on remotely sensed images of different sensors, resolutions and complexity levels. The results were satisfactory since the produced primitive objects, were compared with other segmentation algorithms and are capable of providing meaningful objects through a follow up classification step. An integration of Mseg with an expert system environment will provide an integrated object-oriented image classification system.