Crop Classification Using Multi-Temporal RADARSAT Constellation Mission Compact Polarimetry SAR Data

The RADARSAT Constellation Mission (RCM) Compact Polarimetry (CP) data has become a key asset in crop mapping and monitoring for diverse agricultural landscapes. This study utilizes the unique capabilities of the RCM CP data for crop mapping. It performs a detailed comparison between single-date and...

Full description

Saved in:
Bibliographic Details
Main Authors: Ramin Farhadiani, Saeid Homayouni, Avik Bhattacharya, Masoud Mahdianpari
Format: Article
Language:English
Published: Taylor & Francis Group 2024-12-01
Series:Canadian Journal of Remote Sensing
Online Access:http://dx.doi.org/10.1080/07038992.2024.2384883
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The RADARSAT Constellation Mission (RCM) Compact Polarimetry (CP) data has become a key asset in crop mapping and monitoring for diverse agricultural landscapes. This study utilizes the unique capabilities of the RCM CP data for crop mapping. It performs a detailed comparison between single-date and multi-date classification to underscore the prowess of multi-temporal CP data in crop mapping. The novelty of our approach is in the thorough investigation of real CP data, a significant advancement from previous studies that mainly relied on simulated CP data. The CP data used in this study were acquired on July 1, July 30, and August 27, 2021, over southern Quebec, Canada, including soy, corn, hay, and cereal classes. Various features were extracted from the CP data, and the Random Forest classifier was utilized for crop mapping. The experimental results demonstrated the superiority of multi-temporal CP data for crop classification. The Overall Accuracy (OA) for single-date classifications on July 1, July 30, and August 27 were 61.10%, 75.00%, and 86.45%, respectively. In contrast, the multi-date analysis showed a marked increase in OA (91.20%). This substantial improvement underscores the significant benefit of incorporating multi-date CP data, which delivers a robust and precise framework for crop mapping.
ISSN:1712-7971