IMAGE PROCESSING AND CLUSTERING ALGORITHMS FOR FOREST COVER QUANTIFICATION

“Forest cover” refers to the relative land area covered by forests. Anthropological interventions and the subsequent diminishing forest cover, result in environmental degradation, impacting man-nature interactions. Hence, it became the need of the moment to monitor the forest cover to minimize natural perils and promote sustainable development. The present preliminary work focuses on implementing image processing and k- means clustering techniques on satellite imagery to monitor and quantify the forest cover of the Sundarbans delta, existing across India and Bangladesh. Imagebased algorithms relying on characteristic colouration were proposed for analysing the percentage of forest cover in the predefined area. Among various methods of monitoring and examining forest land, image-based algorithms can be of vital use due to the rise in the accessibility of information and the potential of analysing large data sets with the least processing time. The above-discussed techniques, along with the availability of Machine Learning (ML) and spaceborne photography, will have a futuristic impact on interpreting the variations in land cover and land utilization. Building upon the following algorithm, it is now conceivable to conduct timely comprehensive analysis, real-time evaluation, monitoring, and control on how events unfold. Similarly, data collected from various geographical observation systems may provide several other qualitative features that are more focused.