Image processing for pancake ice detection and size distribution computation

• Published in: International Journal of Remote Sensing
• Main Authors: Parmiggiani, F., Moctezuma-Flores, M., Wadhams, P., Aulicino, G.
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2019
• Subjects: Earth and Planetary Sciences (all); Beaufort Sea; Sea ice
• Online Access: http://hdl.handle.net/11367/78067; https://doi.org/10.1080/01431161.2018.1541367; https://www.tandfonline.com/loi/tres20

This paper presents a processing scheme whose aim is to provide a tool for a rapid measurement of pancake ice size distribution from aerial photographs. The test images used in this study were collected during the flights of the Twin Otter of the Naval Research Laboratory (NRL) which assisted the cruise of the research ship ‘Sikuliaq’ in carrying out an extensive study of autumn sea ice in the southern Beaufort Sea in 2015. The processing scheme is composed of the following steps: i) image enhancement, ii) nonlinear support vector machine (SVM) analysis, iii) marker-controlled watershed segmentation, and iv) ice size distribution computation. The results demonstrate the usefulness of having immediate information on pancake ice size distribution for the subsequent tasks of the field campaign.