Deep Distillation Recursive Network for Remote Sensing Imagery Super-Resolution

Deep convolutional neural networks (CNNs) have been widely used and achieved state-of-the-art performance in many image or video processing and analysis tasks. In particular, for image super-resolution (SR) processing, previous CNN-based methods have led to significant improvements, when compared with shallow learning-based methods. However, previous CNN-based algorithms with simple direct or skip connections are of poor performance when applied to remote sensing satellite images SR. In this study, a simple but effective CNN framework, namely deep distillation recursive network (DDRN), is presented for video satellite image SR. DDRN includes a group of ultra-dense residual blocks (UDB), a multi-scale purification unit (MSPU), and a reconstruction module. In particular, through the addition of rich interactive links in and between multiple-path units in each UDB, features extracted from multiple parallel convolution layers can be shared effectively. Compared with classical dense-connection-based models, DDRN possesses the following main properties. (1) DDRN contains more linking nodes with the same convolution layers. (2) A distillation and compensation mechanism, which performs feature distillation and compensation in different stages of the network, is also constructed. In particular, the high-frequency components lost during information propagation can be compensated in MSPU. (3) The final SR image can benefit from the feature maps extracted from UDB and the compensated components obtained from MSPU. Experiments on Kaggle Open Source Dataset and Jilin-1 video satellite images illustrate that DDRN outperforms the conventional CNN-based baselines and some state-of-the-art feature extraction approaches.

Download Full-text

Deep Residual Squeeze and Excitation Network for Remote Sensing Image Super-Resolution

Remote Sensing ◽

10.3390/rs11151817 ◽

2019 ◽

Vol 11 (15) ◽

pp. 1817 ◽

Cited By ~ 8

Author(s):

Jun Gu ◽

Xian Sun ◽

Yue Zhang ◽

Kun Fu ◽

Lei Wang

Keyword(s):

Remote Sensing ◽

State Of The Art ◽

Super Resolution ◽

Remote Sensing Image ◽

Single Image ◽

Remote Sensing Images ◽

Deep Convolutional Neural Networks ◽

Current Block ◽

Image Super Resolution ◽

Single Image Super Resolution

Recently, deep convolutional neural networks (DCNN) have obtained promising results in single image super-resolution (SISR) of remote sensing images. Due to the high complexity of remote sensing image distribution, most of the existing methods are not good enough for remote sensing image super-resolution. Enhancing the representation ability of the network is one of the critical factors to improve remote sensing image super-resolution performance. To address this problem, we propose a new SISR algorithm called a Deep Residual Squeeze and Excitation Network (DRSEN). Specifically, we propose a residual squeeze and excitation block (RSEB) as a building block in DRSEN. The RSEB fuses the input and its internal features of current block, and models the interdependencies and relationships between channels to enhance the representation power. At the same time, we improve the up-sampling module and the global residual pathway in the network to reduce the parameters of the network. Experiments on two public remote sensing datasets (UC Merced and NWPU-RESISC45) show that our DRSEN achieves better accuracy and visual improvements against most state-of-the-art methods. The DRSEN is beneficial for the progress in the remote sensing images super-resolution field.

Download Full-text

Satellite Image Super-Resolution via Multi-Scale Residual Deep Neural Network

Remote Sensing ◽

10.3390/rs11131588 ◽

2019 ◽

Vol 11 (13) ◽

pp. 1588 ◽

Cited By ~ 13

Author(s):

Tao Lu ◽

Jiaming Wang ◽

Yanduo Zhang ◽

Zhongyuan Wang ◽

Junjun Jiang

Keyword(s):

Neural Network ◽

Remote Sensing ◽

High Frequency ◽

Satellite Images ◽

Satellite Image ◽

Super Resolution ◽

Object Identification ◽

Remote Sensing Images ◽

Frequency Information ◽

Multi Scale

Recently, the application of satellite remote sensing images is becoming increasingly popular, but the observed images from satellite sensors are frequently in low-resolution (LR). Thus, they cannot fully meet the requirements of object identification and analysis. To utilize the multi-scale characteristics of objects fully in remote sensing images, this paper presents a multi-scale residual neural network (MRNN). MRNN adopts the multi-scale nature of satellite images to reconstruct high-frequency information accurately for super-resolution (SR) satellite imagery. Different sizes of patches from LR satellite images are initially extracted to fit different scale of objects. Large-, middle-, and small-scale deep residual neural networks are designed to simulate differently sized receptive fields for acquiring relative global, contextual, and local information for prior representation. Then, a fusion network is used to refine different scales of information. MRNN fuses the complementary high-frequency information from differently scaled networks to reconstruct the desired high-resolution satellite object image, which is in line with human visual experience (“look in multi-scale to see better”). Experimental results on the SpaceNet satellite image and NWPU-RESISC45 databases show that the proposed approach outperformed several state-of-the-art SR algorithms in terms of objective and subjective image qualities.

Download Full-text

Bidirectional Convolutional LSTM Neural Network for Remote Sensing Image Super-Resolution

Remote Sensing ◽

10.3390/rs11202333 ◽

2019 ◽

Vol 11 (20) ◽

pp. 2333 ◽

Cited By ~ 4

Author(s):

Yunpeng Chang ◽

Bin Luo

Keyword(s):

Remote Sensing ◽

Satellite Images ◽

Receptive Fields ◽

Super Resolution ◽

Model Parameters ◽

Deep Convolutional Neural Networks ◽

Additional Information ◽

Image Super Resolution ◽

Block Based ◽

Convolutional Lstm

Single-image super-resolution (SR) is an effective approach to enhance spatial resolution for numerous applications such as object detection and classification when the resolution of sensors is limited. Although deep convolutional neural networks (CNNs) proposed for this purpose in recent years have outperformed relatively shallow models, enormous parameters bring the risk of overfitting. In addition, due to the different scale of objects in images, the hierarchical features of deep CNN contain additional information for SR tasks, while most CNN models have not fully utilized these features. In this paper, we proposed a deep yet concise network to address these problems. Our network consists of two main structures: (1) recursive inference block based on dense connection reuse of local low-level features, and recursive learning is applied to control the model parameters while increasing the receptive fields; (2) a bidirectional convolutional LSTM (BiConvLSTM) layer is introduced to learn the correlations of features from each recursion and adaptively select the complementary information for the reconstruction layer. Experiments on multispectral satellite images, panchromatic satellite images, and nature high-resolution remote-sensing images showed that our proposed model outperformed state-of-the-art methods while utilizing fewer parameters, and ablation studies demonstrated the effectiveness of a BiConvLSTM layer for an image SR task.

Download Full-text

SUPER-RESOLUTION OF MULTISPECTRAL SATELLITE IMAGES USING CONVOLUTIONAL NEURAL NETWORKS

ISPRS Annals of Photogrammetry Remote Sensing and Spatial Information Sciences ◽

10.5194/isprs-annals-v-1-2020-33-2020 ◽

2020 ◽

Vol V-1-2020 ◽

pp. 33-40

Author(s):

M. U. Müller ◽

N. Ekhtiari ◽

R. M. Almeida ◽

C. Rieke

Keyword(s):

Remote Sensing ◽

Neural Networks ◽

Convolutional Neural Networks ◽

Spatial Resolution ◽

Satellite Images ◽

State Of The Art ◽

Super Resolution ◽

Image Resolution ◽

The Other ◽

Multispectral Satellite Images

Abstract. Super-resolution aims at increasing image resolution by algorithmic means and has progressed over the recent years due to advances in the fields of computer vision and deep learning. Convolutional Neural Networks based on a variety of architectures have been applied to the problem, e.g. autoencoders and residual networks. While most research focuses on the processing of photographs consisting only of RGB color channels, little work can be found concentrating on multi-band, analytic satellite imagery. Satellite images often include a panchromatic band, which has higher spatial resolution but lower spectral resolution than the other bands. In the field of remote sensing, there is a long tradition of applying pan-sharpening to satellite images, i.e. bringing the multispectral bands to the higher spatial resolution by merging them with the panchromatic band. To our knowledge there are so far no approaches to super-resolution which take advantage of the panchromatic band. In this paper we propose a method to train state-of-the-art CNNs using pairs of lower-resolution multispectral and high-resolution pan-sharpened image tiles in order to create super-resolved analytic images. The derived quality metrics show that the method improves information content of the processed images. We compare the results created by four CNN architectures, with RedNet30 performing best.

Download Full-text

PCAN—Part-Based Context Attention Network for Thermal Power Plant Detection in Remote Sensing Imagery

Remote Sensing ◽

10.3390/rs13071243 ◽

2021 ◽

Vol 13 (7) ◽

pp. 1243

Author(s):

Wenxin Yin ◽

Wenhui Diao ◽

Peijin Wang ◽

Xin Gao ◽

Ya Li ◽

...

Keyword(s):

Remote Sensing ◽

Power Plants ◽

State Of The Art ◽

Thermal Power ◽

Image Interpretation ◽

Remote Sensing Image ◽

Thermal Power Plants ◽

Average Precision ◽

Deep Convolutional Neural Networks ◽

Multi Scale

The detection of Thermal Power Plants (TPPs) is a meaningful task for remote sensing image interpretation. It is a challenging task, because as facility objects TPPs are composed of various distinctive and irregular components. In this paper, we propose a novel end-to-end detection framework for TPPs based on deep convolutional neural networks. Specifically, based on the RetinaNet one-stage detector, a context attention multi-scale feature extraction network is proposed to fuse global spatial attention to strengthen the ability in representing irregular objects. In addition, we design a part-based attention module to adapt to TPPs containing distinctive components. Experiments show that the proposed method outperforms the state-of-the-art methods and can achieve 68.15% mean average precision.

Download Full-text

Application of Polarization Remote Sensing to Identify Activity of Yitong-Shulan Fault, Jilin Province, China

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.1065-1069.2246 ◽

2014 ◽

Vol 1065-1069 ◽

pp. 2246-2250

Author(s):

Jian Sheng ◽

Guang Yuan Yu ◽

Yu Meng Wang ◽

Han Lv

Keyword(s):

Remote Sensing ◽

Satellite Images ◽

Eastern China ◽

Satellite Image ◽

Jilin Province ◽

Geophysical Method ◽

Fault Activity ◽

Shitoukoumen Reservoir ◽

Observation Method ◽

Buried Fault

Yitong-Shulan fault, one north section of the famed Tanlu grand fault zone in eastern China, is NNE-trending though the Jilin Province, China. In October 2010, Heilongjiang segment of this fault was discovered the evidence of its activity in Holonce, and further inferred it is associated with a paleoearthquake event. So the recognize of Yitong-Shulan fault Jilin section active in the early Quaternary capable of generating moderate quakes is doubted. Yitong-Shulan fault is almost covered by Quaternary strata in Jilin Province. Traditional method is difficult to explore buried fault, and geophysical method is partial and expensive. The polarization remote sensing is a kind of emerging earth observation method, which has high terrain-recognization resolution. The polarization remote sensing method can to indentify the scarps and displaced geomorphic objects along the fault though satellite images. It even can to discover the high of scarps, displacement of geomorphic objects, and so on. The fault activity can be indicated well by the interpretation of polarization remote sensing. In this paper, use the polarization remote sensing method to study the activity of Yitong-Shulan fault Jilin section. Satellite image near the Shulan City, Jilin Province interpreted by polarization remote sensing reveals that the obviously linear scarps which extend long the fault is 1-3m high. Along the fault various kinds of geomorphic objects are displaced. This interpretation result indicated the Shulan-Shitoukoumen Reservoir segment of the fault is active since Holocene. The fault activity also is proved by geophysical method.

Download Full-text

Estimation of PM10 Distribution using Landsat5 and Landsat8 Remote Sensing

Proceedings ◽

10.3390/proceedings2231430 ◽

2018 ◽

Vol 2 (23) ◽

pp. 1430

Author(s):

V. M. Fernández-Pacheco ◽

C. A. López-Sánchez ◽

E. Álvarez-Álvarez ◽

M. J. Suárez López ◽

L. García-Expósito ◽

...

Keyword(s):

Remote Sensing ◽

Air Pollutants ◽

Satellite Images ◽

Satellite Image ◽

Remote Sensing Image ◽

Regional Studies ◽

Landsat 8 ◽

Landsat 8 Oli ◽

Landsat 5 Tm ◽

Source Of Information

Air pollution is one of the major environmental problems, especially in industrial and highly populated areas. Remote sensing image is a rich source of information with many uses. This paper is focused on estimation of air pollutants using Landsat-5 TM and Landsat-8 OLI satellite images. Particulate Matter with particle size less than 10 microns (PM10) is estimated for the study area of Principado de Asturias (Spain). When a satellite records the radiance of the surface received at sensor, does not represent the true radiance of the surface. A noise caused by Aerosol and Particulate Matters attenuate that radiance. In many applications of remote sensing, that noise called path radiance is removed during pre-processing. Instead, path radiance was used to estimate the PM10 concentration in the air. A relationship between the path radiance and PM10 measurements from ground stations has been established using Random Forest (RF) algorithm and a PM10 map was generated for the study area. The results show that PM10 estimation through satellite image is an efficient technique and it is suitable for local and regional studies.

Download Full-text

Transferred Multi-Perception Attention Networks for Remote Sensing Image Super-Resolution

Remote Sensing ◽

10.3390/rs11232857 ◽

2019 ◽

Vol 11 (23) ◽

pp. 2857 ◽

Cited By ~ 3

Author(s):

Xiaoyu Dong ◽

Zhihong Xi ◽

Xu Sun ◽

Lianru Gao

Keyword(s):

Remote Sensing ◽

Learning Strategy ◽

State Of The Art ◽

Spatial Scales ◽

Super Resolution ◽

Natural Image ◽

Training Procedure ◽

Remote Sensing Images ◽

Objective Criterion ◽

Image Super Resolution

Image super-resolution (SR) reconstruction plays a key role in coping with the increasing demand on remote sensing imaging applications with high spatial resolution requirements. Though many SR methods have been proposed over the last few years, further research is needed to improve SR processes with regard to the complex spatial distribution of the remote sensing images and the diverse spatial scales of ground objects. In this paper, a novel multi-perception attention network (MPSR) is developed with performance exceeding those of many existing state-of-the-art models. By incorporating the proposed enhanced residual block (ERB) and residual channel attention group (RCAG), MPSR can super-resolve low-resolution remote sensing images via multi-perception learning and multi-level information adaptive weighted fusion. Moreover, a pre-train and transfer learning strategy is introduced, which improved the SR performance and stabilized the training procedure. Experimental comparisons are conducted using 13 state-of-the-art methods over a remote sensing dataset and benchmark natural image sets. The proposed model proved its excellence in both objective criterion and subjective perspective.

Download Full-text

Contenido y calidad de las imágenes de observación terrestre

Revista ECIPeru ◽

10.33017/reveciperu2012.0015/ ◽

2019 ◽

pp. 15-21

Keyword(s):

Remote Sensing ◽

Image Quality ◽

Information Extraction ◽

Satellite Images ◽

Satellite Image ◽

La Paz ◽

Sensing System ◽

Manual Intervention ◽

Shape Structure ◽

Important Amount

Contenido y calidad de las imágenes de observación terrestre Earth observation image information content and quality Avid Roman-Gonzalez, Natalia Indira Vargas-Cuentas TELECOM ParisTech, 46 rue Barrault, 75013 – Paris, Francia Escuela Militar de Ingeniería – EMI, La Paz, Bolivia DOI: https://doi.org/10.33017/RevECIPeru2012.0015/ Resumen En el presente artículo describiremos la extracción de información de imágenes satelitales y la importancia de la calidad de las imágenes satelitales. Indagaremos con más detalle en el ámbito de los artefactos y su influencia en la extracción de información de las imágenes satelitales. En un sistema de teledetección, si bien, las imágenes son muy importantes, pero lo más importante es la información que podemos extraer de ellas para interpretar y aplicar esta información en diferentes campos. En ese sentido, la calidad de imagen juega un papel importante. Si queremos obtener la mayor e importante cantidad de información de una imagen, es necesario que la imagen tenga una buena calidad. El principal objetivo de cualquier sistema de teledetección es el uso de la información que se puede extraer de las imágenes, esto incluye la detección, medición, identificación e interpretación de diferentes objetivos de interés. Los objetivos de interés en imágenes de teledetección pueden ser cualquier característica, objeto, textura, forma, estructura, espectro o cobertura superficial que están en la imagen. El proceso de un sistema de teledetección y análisis puede ser realizado manualmente o de manera automática, en realidad, hay muchos grupos de investigación que desarrollan diferentes herramientas para detectar, identificar, interpretar y extraer información de los objetivos de interés sin intervención manual de un intérprete humano. Descriptores: teledetección, imágenes satelitales, detección de artefactos, calidad de las imágenes. Abstract In this article we will describe the information extraction from satellite image, the importance of image quality in satellite image. In this paper we will study in more detail the artifacts and their influence on the information extraction from satellite images. In a remote sensing system, although, the images are very important, but more important is the information that we can extract from them to interpret and apply this information in different fields. In this sense, the image quality plays an important role. If we want to get the biggest and most important amount of information from the image, we need to have a good image quality. The main objective of any remote sensing system is the use of information that we can extract from the images, this includes detection, measurement, identification and interpretation of different targets. Targets in remote sensing images may be any feature, object, texture, shape, structure, spectrum or land covers which are in the image. Remote sensing process and analysis could be performed manually or automatically, actually, there are many research groups that develop different tools for detect, identify, extract information and interpret targets without manual intervention by a human interpreter. Keywords: remote sensing, satellite images, artifacts detection, image quality.

Download Full-text

Estimation of Forest Stand Parameters Using SPOT-5 Satellite Images and Topographic Information

10.20944/preprints201710.0017.v1 ◽

2017 ◽

Author(s):

Shiqin Xie ◽

Wei Wang ◽

Qian liu ◽

Jinghui Meng ◽

Tianzhong Zhao ◽

...

Keyword(s):

Remote Sensing ◽

Satellite Images ◽

Satellite Image ◽

Forest Stand ◽

Textural Features ◽

Spectral Features ◽

Topographic Information ◽

Study Results ◽

Stand Parameters ◽

Spot 5

In recent years, remote sensing technology has been widely used to predict forest stand parameters. In order to compare the effects of different features of remote sensing images and topographic information on the prediction of forest stand parameters, multivariate stepwise regression analysis method was used to build estimation models for important forest stand parameters by using textural and spectral features as well as topographic information of SPOT-5 satellite images in northeastern Heilongjiang Province in China as independent variables. The study results show that the optimal window to predict forest stand parameters using textural features of SPOT-5 satellite image is 9×9; the ability of textural features was better than that of spectral features in terms of predicting forest stand parameters; with the inclusion of topographic information, the accuracy of prediction of all models was improved, of which elevation has the most significant effect. The highest accuracy was achieved when predicting the stand volume (SV) (R2adj=0.820), followed by basal area (BA) (R2adj =0.778), accuracy of both above models exceeded 75%. The results show that models combined use of textural, spectral features and topographic information of SPOT-5 images have a good application prospect in predicting forest stand parameters.

Download Full-text