No abstract

The unique climate and topography of the Tibetan Plateau produce an abundant distribution of lakes. These lakes are important indicators of climate change, and changes in lake area have critical implications for water resources and ecological conditions. Lake area change can be monitored using the huge sets of high-resolution remote sensing data available, but this demands an automatic water classification system. This study develops an algorithm for automatic water classification using Chinese <small>GF</small>-1 (or Gaofen-1) wide-field-of-view (<small>WFV</small>) satellite data. The original <small>GF</small>-1 <small>WFV</small> data were automatically preprocessed with radiometric correction and orthorectification. The single-band threshold and two global-local segmentation methods were employed to distinguish water from non-water features. Three methods of determining the optimal thresholds for normalized difference water index (<small>NDWI</small>) images were compared: Iterative Self Organizing Data Analysis Technique (<small>ISODATA</small>); global-local segmentation with thresholds specified by stepwise iteration; and the Otsu method. The water classification from two steps of global-local segmentations showed better performance than the single-band threshold and <small>ISODATA</small> methods. The <small>GF</small>-1 <small>WFV</small>-based lake mapping across the entire Tibetan Plateau in 2015 using the global-local segmentations with thresholds from the Otsu method showed high quality and efficiency in automatic water classification. This method can be extended to other satellite datasets, and makes the high-resolution global monitoring and mapping of lakes possible.

Highlights A novel weakly supervised learning framework for COVID-19 severity assessment A multiple instance learning model with virtual bag-based augmentation A novel self-supervised pretext task to aid the learning process

No abstract

Complex wavelet structural similarity (CW-SSIM) index has been proposed as a promising image similarity measure that is robust to small geometric distortions such as translation, scaling and rotation of images, but how to make the best use of it in image classification problems has not been deeply investigated. In this paper, we propose a novel "feature-extraction free" image classification algorithm based on CW-SSIM and use handwritten digit recognition as an example to demonstrate it. First, a CW-SSIM based unsupervised clustering method is used to divide the training images into clusters and to pick a representative image for each cluster. A supervised learning method based on support vector machines is then employed to maximize the classification accuracy based on CW-SSIM values between an input image and the representative images. Our experiments show that such a conceptually simple image classification method, which does not involve any registration, intensity normalization or sophisticated feature extraction processes, and does not rely on any modeling of the image patterns or distortion processes, achieves competitive performance with reduced computational complexity.