Land cover (LC) mapping is essential for monitoring the environment and understanding the effects of human activities on it. Recent studies demonstrated successful applications of specific deep learning models to small-scale LC mapping tasks (e.g., wetland mapping). However, it is not readily clear which of the existing state-of-the-art models for natural images are the best candidates to be taken for the particular remote sensing task and data. In this study, we answer that question for mapping the fundamental LC classes using the satellite imaging radar data. We took ESA Sentinel-1 C-band SAR images acquired during the whole summer season of 2018 in Finland, which are representative of the land cover in the country. CORINE LC map was used as a reference, and the models were trained to distinguish between the 5 major CORINE based classes. We selected seven among the state-of-the-art semantic segmentation models so that they cover a diverse set of approaches: U-Net, DeepLabV3+, PSPNet, BiSeNet, SegNet, FC-DenseNet, and FRRN-B, and further fine-tuned them. Upon evaluation and benchmarking, all the models demonstrated solid performance with overall accuracy between 87.9% and 93.1%, with good to a very good agreement (kappa statistic between 0.75 and 0.86). The two best models were FC-DenseNet (Fully Convolutional DenseNets) and SegNet (Encoder-Decoder-Skip), with the latter having a much shorter inference time. Overall, our results indicate that the semantic segmentation models are suitable for efficient wide-area mapping using satellite SAR imagery and provide baseline accuracy against which the newly proposed models should be evaluated.

The ICEYE constellation features the first operational microsatellite based X-band SAR sensors suitable for allweather day-and-night Earth Observation. In this paper we report on the status of the ICEYE Constellation and describe the characteristics of the first operational imaging modes.

No abstract

In 2015, ICEYE, a Finnish radar satellite company, with support by ExxonMobil Upstream Research Company, executed a series of field tests to assess the technical feasibility of using ICEYE's newly-designed Synthetic Aperture Radar (SAR) instrument for small ice feature detection in open water and pack ice imaging. The project consisted of three separate flight campaigns. The first campaign included calibration flights flown in Helsinki in March. The second campaign was performed in April, over sea ice in the Gulf of Bothnia. The third flight campaign, conducted in November, measured the performance of the instrument in weather. In addition to validation of the designed SAR system, the main objective of the project was to obtain imagery of small ice features with the ICEYE prototype instrument. The analysis focused on the capability of the proposed SAR to detect small but potentially hazardous ice features. Two instrument configurations were used: a linear polarized antenna in VV configuration, and a circular cross-polarized antenna configuration. The reason for using a circular cross-polarized configuration in the SAR is that potential rain clutter can be reduced by suppressing odd-numbered reflections. Test results from the ice measurement campaign showed that the linear VV configuration is sufficient for detecting features as small as 10 meters in waterline extension. On the other hand, small ice feature detection performance of the circular cross-polarized configuration was poor at transmitted power levels scaled to match the satellite case. Results also showed that sea state and target feature roughness play a significant role in detecting small features. The third campaign was performed to understand whether rain clutter would affect the image quality with VV configuration, and whether such effects can be mitigated by a circular cross-polarized configuration. However, during the test period, only light to moderate rain conditions were encountered. In these conditions, the VV configuration suffered no rain clutter in levels affecting ice detection performance. A theoretical study concluded that in order for the rain clutter to interfere with imaging, the rain event must exceed levels that are extremely rare in arctic areas. The flight test demonstrates that the ICEYE SAR data is valuable for ice management use. Once deployed in a constellation of small satellites, the resulting near-real time service can provide better situational awareness for potential operations in the Arctic.