A Multiview Semantic Vegetation Index for Robust Estimation of Urban Vegetation Cover

Urban vegetation growth is vital for developing sustainable and liveable cities in the contemporary era since it directly helps people’s health well-being. Estimating cover biomass commonly done by calculating various indices automated urban management monitoring. However, most of these fail to capture robust estimation due their inherent focus on colour attributes with limited viewpoint ignore seasonal changes. To solve this limitation, article proposed a novel index called Multiview Semantic Vegetation Index (MSVI), which color, viewpoint, variations. Moreover, can be applied RGB images. This (MSVI) based deep semantic segmentation multiview field coverage integrated into any platform. has been tested Google Street View (GSV) imagery Wyndham City Council, Melbourne, Australia. The experiments training achieved an overall pixel accuracy 89.4% 92.4% FCN U-Net, respectively. Thus, MSVI helpful instrument analysing forestry provides accurate reliable objective method assessing plant at street level.