The K means clustering was processed for threshold vegetation indices and gap detection. It was processed for retrieving the vegetation index value that represents forest land cover, percentage vegetation coverage, and canopy density. The method was further used for finding the probability distribution of forest canopy gaps in the forest. The result was tested in the Hazaribagh Wildlife Sanctuary, Jharkhand, India. The percentage vegetation cover was calculated in the new SNAP software. The canopy density was mapped through FCD model. From the analysis, it was estimated that the dense forest having greater than 70% of canopy density comprises 64-100% of vegetation cover; moderately dense forest having 40-70% canopy density includes 21-64% of vegetation cover, and open forest having less than 40% canopy density have 7-21% of vegetation cover. The Normalized Vegetation Index (NDVI) and Transformed Vegetation Index (TVI) considered being more efficient and Difference Vegetation Index (DVI) was less efficient for forest vegetation cover and density measurement. Inversely, it was observed that DVI was more efficient in finding gaps in the forest. The method was also functional for finding the probability distribution of canopy gaps in the forest. This clustering technique can be applied in other means for forest landscape level assessment.