Modified Convolutional Neural Network Based on Dropout and the Stochastic Gradient Descent Optimizer

Fan

et al. 2018

Water

178

Accurate information on urban surface water is important for assessing the role it plays in urban ecosystem services in the context of human survival and climate change. The precise extraction of urban water bodies from images is of great significance for urban planning and socioeconomic development. In this paper, a novel deep-learning architecture is proposed for the extraction of urban water bodies from high-resolution remote sensing (HRRS) imagery. First, an adaptive simple linear iterative clustering algorithm is applied for segmentation of the remote-sensing image into high-quality superpixels. Then, a new convolutional neural network (CNN) architecture is designed that can extract useful high-level features of water bodies from input data in a complex urban background and mark the superpixel as one of two classes: an including water or no-water pixel. Finally, a high-resolution image of water-extracted superpixels is generated. Experimental results show that the proposed method achieved higher accuracy for water extraction from the high-resolution remote-sensing images than traditional approaches, and the average overall accuracy is 99.14%.

Section: Self-adaptive Pooling Convolutional Neural Network (Cnn) Armentioning

confidence: 99%

Extraction of Urban Water Bodies from High-Resolution Remote-Sensing Imagery Using Deep Learning

Fan

et al. 2018

Water

178

“…The α parameter is used to control the relative importance of color similarity and spatial proximity. The attraction function reflects the possibility of the jth pixel attracting the ith pixel as its cluster [25]. The attraction function is expressed as:…”

Section: Preprocessingmentioning

confidence: 99%

Multilevel Cloud Detection for High-Resolution Remote Sensing Imagery Using Multiple Convolutional Neural Networks

Fan

Bilal

et al. 2018

IJGI

In high-resolution image data, multilevel cloud detection is a key task for remote sensing data processing. Generally, it is difficult to obtain high accuracy for multilevel cloud detection when using satellite imagery which only contains visible and near-infrared spectral bands. So, multilevel cloud detection for high-resolution remote sensing imagery is challenging. In this paper, a new multilevel cloud detection technique is proposed based on the multiple convolutional neural networks for high-resolution remote sensing imagery. In order to avoid input the entire image into the network for cloud detection, the adaptive simple linear iterative clustering (A-SCLI) algorithm was applied to the segmentation of the satellite image to obtain good-quality superpixels. After that, a new multiple convolutional neural networks (MCNNs) architecture is designed to extract multiscale features from each superpixel, and the superpixels are marked as thin cloud, thick cloud, cloud shadow, and non-cloud. The results suggest that the proposed method can detect multilevel clouds and obtain a high accuracy for high-resolution remote sensing imagery.

“…After 8800 iterations, the test error of CNN, CNN with L 2 and PCNN reaches 2.84%, 2.77%, and 2.65%,respectively. [28] 99.22 Graph − CNN [29] 99.14 CNN MA [30] 98.75 MCNN − DS [31] 98.43…”

Section: Mnistmentioning

confidence: 99%

A New Improved Learning Algorithm for Convolutional Neural Networks

Zhao

et al. 2020

Processes

The back-propagation (BP) algorithm is usually used to train convolutional neural networks (CNNs) and has made greater progress in image classification. It updates weights with the gradient descent, and the farther the sample is from the target, the greater the contribution of it to the weight change. However, the influence of samples classified correctly but that are close to the classification boundary is diminished. This paper defines the classification confidence as the degree to which a sample belongs to its correct category, and divides samples of each category into dangerous and safe according to a dynamic classification confidence threshold. Then a new learning algorithm is presented to penalize the loss function with danger samples but not all samples to enable CNN to pay more attention to danger samples and to learn effective information more accurately. The experiment results, carried out on the MNIST dataset and three sub-datasets of CIFAR-10, showed that for the MNIST dataset, the accuracy of Non-improve CNN reached 99.246%, while that of PCNN reached 99.3%; for three sub-datasets of CIFAR-10, the accuracies of Non-improve CNN are 96.15%, 88.93%, and 94.92%, respectively, while those of PCNN are 96.44%, 89.37%, and 95.22%, respectively.