D. Singh, S. Mondal, R. S. Hooda
2018 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
<p><strong>Abstract.</strong> Urban Green Space (UGS) enhances the structure of the landscape of a city. It provides economic, social and health benefits to the inhabitants. Rapid urbanization and increase in population has put extreme pressure on urban green spaces. Therefore, mapping of urban green spaces is important to facilitate the urban planning sustainably. This study has been taken up to map the UGS in the Hisar Municipal Corporation (HMC) area of Haryana State constituting an area of
more » ... ituting an area of 4197.32<span class="thinspace"></span>ha using high resolution LISS-IV data of November, 2017. Three different methods namely (a) Onscreen Digitization (b) Support Vector Machine-based classification (SVM) and (c) Multi Resolution Segmentation (MRS) have been tested to extract the green spaces. The onscreen digitization method has been considered as best method as compared to SVM and MRS because of the intervention of human mind and full control on manual editing. The SVM under estimated the green space for HMC with average relative deviation of 15%. However, it extracted the green area precisely without mixing of open spaces. MRS have given good results in terms of total area (relative deviation was 5%), however, spatial distribution of green spaces have been mixed with open spaces. The study identified the green index of HMC and suggested the suitability and sustainability of the city based on World Health Organization (WHO) defined standards. Green space for Hisar Municipal Corporation was 1690.96<span class="thinspace"></span>ha with average urban green space index of 0.40. Per capita green cover for HMC was 9.69<span class="thinspace"></span>m<sup>2</sup>/inhabitant, which is as per the standard reported by WHO i.e. 9<span class="thinspace"></span>m<sup>2</sup>/inhabitant. Further study may suggest some other algorithm for better results and automatic green space extraction.</p>
doi:10.5194/isprs-archives-xlii-5-921-2018 fatcat:2fyzrffzujhylnq3opp57kcjj4