Machine learning on high performance computing for urban greenspace change detection: satellite image data fusion approach. (2nd July 2020)

• Record Type: Journal Article
• Authors: More, Nilkamal; Nikam, V. B.; Banerjee, Biplab
• Is Part Of: International journal of image and data fusion. Volume 11:Number 3(2020)
• Page Start: 218
• Page End: 232
• Publication Date: 2020-07-02
• Subjects: Satellite image analysis -- data fusion -- SVM classifier -- GPU computing -- urban change analysis; Image processing -- Periodicals; Multisensor data fusion -- Periodicals; Multisensor data fusion; Periodicals; 621.36705
• DOI: 10.1080/19479832.2020.1749142
• Languages: English
• ISSNs: 1947-9832

Abstract

ABSTRACT: Green spaces serve important environmental and quality-of-life functions in urban environments. Fast-changing urban regions require continuous and fast green space change detection. This study focuses on assessment of green space change detection using GPU- for time efficient green space identification and monitoring. Using spatio-temporal data from satellite images and a support vector machine (SVM) as a classification algorithm, this research proposes a platform for green space analysis and change detection. The main contributions of this research include the fusion of the thermal band in addition to Near infra-red, red, green band with the fusion of high spectral information of the moderate resolution imaging spectroradiometer (MODIS) dataset and high spatial information of the LANDSAT 7 dataset. The novel method is employed to calculate the total green space area in the Mumbai metropolitan area and monitor the changes from 2005–2019. This research paper discusses the findings of our strategy and reveals that over the course of 15 years the overall green space was reduced to 50%.