Abstract:

Objective- A satellite images of the earth (or geospatial images) are critical sources of information in diverse fields such as geography, cartography, meteorology, surveillance, city planning. These images contain visual information about various natural and man-made features on or above the surface of the earth. Manual annotation of geospatial images covering even a relatively small area of the earth is a tedious task. This has necessitated research into automated annotation of geospatial images.

Design / Methodology/ Approach-   An important component of this research comprises object detection methods, which are model-driven methods that seek to identify probable locations of specified features of interest or objects in geospatial images.

Findings- High Resolution remote sensing images offer a more detailed description -of the observed scene. However, most of these objects could be complex structures and surrounded by disturbing background, which make object detection and image interpretation even more difficult.

Limitations- to investigate more variations on the basic energy function to produce superpixels with other interesting properties, such as certain predetermined orientations, etc. Another direction is to change patch size as a function of local image variance. It can also use our algorithm to integrate results from different segmentation algorithms.

Practical implications- This paper inspires research scholars, industrialist and academicians who are related to geospatial objects.

Originality/Value- This paper contains different types of models used in segmentation and classification of geospatial objects as literature survey.

Keywords- geospatial images, annotation, Resolution