An Object Similarity-Based Thresholding Method for Urban Area Mapping from Visible Infrared Imaging Radiometer Suite Day/Night Band (VIIRS DNB) Data

• Main Authors: Wenting Ma, Peijun Li
• Format: Article
• Language: English
• Published: MDPI AG 2018-02-01
• Series: Remote Sensing
• Subjects: VIIRS DNB; urban area; threshold; segmentation; similarity
• Online Access: http://www.mdpi.com/2072-4292/10/2/263

Nighttime light data from the Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) provides a unique data source for mapping and monitoring urban areas at regional and global scales. This study proposes an object similarity-based thresholding method using VIIRS DNB data to map urban areas. The threshold for a target potential urban object was determined by comparing its similarity with all reference urban objects with known optimal thresholds derived from Landsat data. The proposed method includes four major steps: potential urban object generation, threshold optimization for reference urban objects, object similarity comparison, and urban area mapping. The proposed method was evaluated using VIIRS DNB data of China and compared with existing mapping methods in terms of threshold estimation and urban area mapping. The results indicated that the proposed method estimated thresholds and mapped urban areas accurately and generally performed better than the cluster-based logistic regression method. The correlation coefficients between the estimated thresholds and the reference thresholds were 0.9201–0.9409 (using Euclidean distance as similarity measure) and 0.9461–0.9523 (using Mahalanobis distance as similarity measure) for the proposed method and 0.9435–0.9503 for the logistic regression method. The average Kappa Coefficients of the urban area maps were 0.58 (Euclidean distance) and 0.57 (Mahalanobis distance) for the proposed method and 0.51 for the logistic regression method. The proposed method shows potential to map urban areas at a regional scale effectively in an economic and convenient way.