Remote sensing is now essential across many fields, thanks to advanced techniques and expanding applications. Some objects may share similar geographical conditions but possess varied spectral properties, while others may differ in geographical features but display similar spectral properties. This illustrates that spectral information alone cannot suffice for precise spatial information, thus emphasizing the significance of spatial and contextual information. Measures of homogeneity and heterogeneity frequently assess image criteria, including spectrum, space, texture, shape, size, context, time, and prior knowledge. Therefore, many researchers have shifted their focus toward unconventional methods like Object-Based Image Analysis (OBIA) to extract data from high-resolution images with greater precision. The first step in the OBIA technique is segmentation, which involves dividing an image into relatively homogeneous areas or segments. Selecting appropriate segmentation parameters compactness, shape, and scale is a fundamental stage in the image segmentation process. There is currently a shortage of global models or frameworks for computing scale parameters, as well as a lack of universal methods or algorithms in this area. It is important to note that there is no one-size-fits-all scale for image objects with varying sizes, shapes, and spatial distributions that are present in a scene.
The main objective of this research is to identify the optimal values for the parameters used in image segmentation. Therefore, this research has utilized Worldview-3, Worldview-2, and GeoEye-1 images with varying parameter values to understand the relationship between parameters and image resolution by keeping most variables fixed and using different-resolution images of the same area.