New Summit website coming in May 2021!

                   Check the SFU library website for updates.

Cloud and Cloud Shadow Segmentation for Remote Sensing Imagery via Filtered Jaccard Loss Function and Parametric Augmentation

Peer reviewed: 
Yes, item is peer reviewed.
Scholarly level: 
Graduate student (PhD)
Final version published as: 

S. Mohajerani, & P. Saeedi. (2021). Cloud and Cloud Shadow Segmentation for Remote Sensing Imagery via Filtered Jaccard Loss Function and Parametric Augmentation. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 1–1. https://doi.org/10.1109/JSTARS.2021.3070786.

Date created: 
2021-04-02
Identifier: 
DOI: 10.1109/JSTARS.2021.3070786
Keywords: 
Clouds
Remote sensing
Image segmentation
Earth
Artificial satellites
Training
Feature extraction
Cloud detection
CNN
Image segmentation
Deep learning
Landsat 8
Loss function
Remote sensing
38-Cloud
Abstract: 

Cloud and cloud shadow segmentation are fundamental processes in optical remote sensing image analysis. Current methods for cloud/shadow identification in geospatial imagery are not as accurate as they should, especially in the presence of snow and haze. This paper presents a deep learning-based framework for the detection of cloud/shadow in Landsat 8 images. Our method benefits from a convolutional neural network, Cloud-Net+ (a modification of our previously proposed Cloud-Net) that is trained with a novel loss function (Filtered Jaccard Loss). The proposed loss function is more sensitive to the absence of foreground objects in an image and penalizes/rewards the predicted mask more accurately than other common loss functions. In addition, a sunlight direction-aware data augmentation technique is developed for the task of cloud shadow detection to extend the generalization ability of the proposed model by expanding existing training sets. The combination of Cloud-Net+, Filtered Jaccard Loss function, and the proposed augmentation algorithm delivers superior results on four public cloud/shadow detection datasets. Our experiments on Pascal VOC dataset exemplifies the applicability and quality of our proposed network and loss function in other computer vision applications.

Language: 
English
Document type: 
Article
File(s): 
Statistics: