DATOS publishes first international publication on automated mapping of sugarcane
The Remote Sensing and Data Science Helpdesk: DATOS Help Desk team, in cooperation with the Sugar Regulatory Administration (SRA), published the paper entitled "Sugarcane Plantation Mapping Using Dynamic Time Warping from Multi-temporal Sentinel-1A Radar Images" on the ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences.
The ISPRS, or the International Society for Photogrammetry and Remote Sensing, is said to be the largest and oldest organization for geomatics practitioners.
Dynamic Time Warping
This paper details the use of Dynamic Time Warping (DTW), originally used for voice recognition, in detecting sugarcane plantations from a series of radar satellite images. Seasonal crops have a fixed duration for its life cycle from planting to harvesting. Sugar cane, for instance, has a 10 to 12-month phenological cycle. This growth pattern allows it to exhibit a unique characteristic in terms of radar backscatter and time.
The researchers, headed by DOST-ASTI Supervising Science Research Specialist Engr. Nestor Olfindo, acquired satellite images of existing sugarcane plantations. From these images, the radar backscatter of these plantations during each stage of the sugarcane’s life cycle was analysed. A temporal signature, which served as an identifier for the life cycle of the sugarcane, was determined. This temporal signature is akin to a voice pattern for this specific crop and was matched with other locations to determine if they were also plantations. The DTW algorithm was used to compare and match the temporal signatures to determine and map out other sugarcane plantations from satellite images.
To verify the detected sugarcane plantations, the DATOS team collaborated with SRA and used the agency’s existing maps and inventory of sugarcane locations. This was formalized in a Memorandum of Agreement signed last 2018 (Related: DATOS holds training series with SRA). The outputs of this partnership are being used by the SRA for their nationwide sugarcane mapping operations. The result aided the agency in updating the inventory of sugarcane plantations faster with detection accuracy of more than 92 percent.
The DTW for sugarcane mapping is part of the DATOS team’s research on the use of the algorithm on satellite images for mapping seasonal crops. A patent for this technology was also successfully filed by the project earlier this year.
The paper will be part of the on-demand catalogue in the 25th ISPRS Congress, which will take place online from 31 August to 02 September 2020. Originally slated to happen in Nice, France, this year's event will now be available virtually to the public. Check out the list of presentations and registration information here: http://www.isprs2020-nice.com/index.php/virtualevent-2.
Two other representatives from the DOST-ASTI will be presenting via a pre-recorded video:
- Roel M. de la Cruz (Near-real time flood detection from multi-temporal sentinel radar images using artificial intelligence)
- Rocell Niño Vicente (Earth observation applications for Goal 14: Improving maritime domain awareness using synthetic aperture radar imaging with automatic identification system in the Philippines).
About the ISPRS
Established in 1910 by Professor Eduard Doležal from the Vienna University of Technology as the International Society for Photogrammetry (ISP), the International Society for Photogrammetry and Remote Sensing (ISPRS) is a non-governmental organisation devoted to the development of international cooperation for the advancement of photogrammetry, remote sensing and spatial information sciences and their applications. The ISPRS Annals contain selected full paper double-blind peer-reviewed scientific contributions of ISPRS Congresses, Symposia and Conferences and Workshops. The series was newly established in 2012.