<![CDATA[Remote sensing with high-resolution panchromatic imagery plays an important role in spatial analysis such as regional mapping, spatial planning, and disaster mitigation. One important stage in image processing is edge detection, which serves to clarify the boundaries of objects such as buildings, roads, and rivers. Problems that often arise in panchromatic images are noise, contrast differences, and spatial data complexity, which cause traditional edge detection methods to be less than optimal. This study proposes the use of the Homogeneity algorithm as a solution due to its simplicity in gradient calculation. This algorithm works by comparing the intensity value of the central pixel to its eight neighbours in a 3x3 kernel, then selecting the maximum difference to determine the presence of edges. The results of implementation on a 5x5 sample image show that large difference values represent edges, while small values indicate homogeneous areas. Further application on a 500x500 pixel panchromatic image produces clear and contrasting edge visualisation, allowing for better recognition of surface object boundaries. This study proves that the Homogeneity algorithm is capable of extracting image edges quickly and effectively, and can be used as an efficient alternative to complex convolution methods for real-time remote sensing applications.]]>
Copyrights © 2025
