Low-resolution Datasets Research Articles

Erosion and deposition vulnerability of small (

The tropics are naturally vulnerable to watershed erosion. This region is rapidly growing (projected to be 50% of the global population by 2050) which exacerbates erosional issues by the subsequent land use change. The issue is particularly of interest on the many (~45,000) small tropical (<5,000 km2) islands, and their >115M residents, where ecotourism and sediment intolerant ecosystems such as coral reefs are the main driver of their economies. However, vulnerability to erosion and deposition is poorly quantified in these regions due to the misclassification or exclusion of small islands in coarse global analyses. We use the only vulnerability assessment method that connects watershed erosion and coastal deposition to compare locally sourced, high-resolution datasets (5 x 5 m) to satellite-collected, remotely sensed low-resolution datasets (463 x 463 m). We find that on the island scale (~52 km2) the difference in vulnerability calculated by the two methods is minor. On the watershed scale however, low-resolution datasets fail to accurately demonstrate watershed and coastal deposition vulnerability when compared to high-resolution analysis. Specifically, we find that anthropogenic development (roads and buildings) is poorly constrained at a global scale. Structures and roads are difficult to identify in heavily forested regions using satellite algorithms and the rapid, ongoing rate of development aggravates the issue. We recommend that end-users of this method obtain locally sourced anthropogenic development datasets for the best results while using low resolution datasets for the other variables. Fortunately, anthropogenic development data can be easily collected using community-based research or identified using satellite imagery by any level of user. Using high-resolution results, we identify a development trend across St. John and regions that are both high risk and possible targets for future development. Previously published modeled and measured sedimentation rates demonstrate the method is accurate when using low-resolution or high-resolution data but, anthropogenic development, watershed slope, and earthquake probability datasets should be of the highest resolution depending on the region specified.

Ultrasound Image Super-Resolution with Two-Stage Zero-Shot CycleGAN

Medical ultrasound imaging is widely used in clinical diagnosis because of its non-invasive, convenient and quick characteristics. However, due to its low image contrast, multiple artifacts, noise and lack of paired high-resolution and low-resolution image data sets, the task of super-resolution reconstruction of medical ultrasound images is more challenging. In this paper, the Two-Stage GAN network model was adjusted by CycleGAN generation and unsupervised learning methods, and the Two-Stage ZSSR (“Zero-Shot” Super-Resolution) CycleGAN network was proposed. The objective evaluation indexes PSNR and SSIM were raised to 40.8079 and 0.9953. The visual effect was also significantly improved.

A Compilation of Snow Cover Datasets for Svalbard: A Multi-Sensor, Multi-Model Study

Reliable and accurate mapping of snow cover are essential in applications such as water resource management, hazard forecasting, calibration and validation of hydrological models and climate impact assessments. Optical remote sensing has been utilized as a tool for snow cover monitoring over the last several decades. However, consistent long-term monitoring of snow cover can be challenging due to differences in spatial resolution and retrieval algorithms of the different generations of satellite-based sensors. Snow models represent a complementary tool to remote sensing for snow cover monitoring, being able to fill in temporal and spatial data gaps where a lack of observations exist. This study utilized three optical remote sensing datasets and two snow models with overlapping periods of data coverage to investigate the similarities and discrepancies in snow cover estimates over Nordenskiöld Land in central Svalbard. High-resolution Sentinel-2 observations were utilized to calibrate a 20-year MODIS snow cover dataset that was subsequently used to correct snow cover fraction estimates made by the lower resolution AVHRR instrument and snow model datasets. A consistent overestimation of snow cover fraction by the lower resolution datasets was found, as well as estimates of the first snow-free day (FSFD) that were, on average, 10–15 days later when compared with the baseline MODIS estimates. Correction of the AVHRR time series produced a significantly slower decadal change in the land-averaged FSFD, indicating that caution should be exercised when interpreting climate-related trends from earlier lower resolution observations. Substantial differences in the dynamic characteristics of snow cover in early autumn were also present between the remote sensing and snow model datasets, which need to be investigated separately. This work demonstrates that the consistency of earlier low spatial resolution snow cover datasets can be improved by using current-day higher resolution datasets.

Remote sensing image super‐resolution based on convolutional blind denoising adaptive dense connection

The current super-resolution (SR) deep network is mainly applied to the common image and pays little attention to the image with noise. The remote sensing image contains much noise, so that the SR reconstruction effect is not satisfactory. Therefore, a convolution blind denoising adaptive dense connection SR (CBD-ADCSR) network for the remote sensing image is proposed in this paper. The whole model is divided into a convolution blind denoising (CBD) network for denoising and an ADCSR network for reconstruction. Firstly, the components of the network are given in detail and are analysed. Secondly, a data set making method is designed combining motion blur, defocusing blur and Gaussian noise, which is used to generate low-resolution image data sets with complex degradation for the model training. Finally, through the detailed comparative experiment, it is proved that the reconstruction effect of the CBD-ADCSR model is better than that of the most state-of-the-art algorithms in objective criteria. In addition, compared with the original ADCSR network, CBD-ADCSR has a stronger ability for noise suppression.

Climate and Socioeconomic Factors Drive Irrigated Agriculture Dynamics in the Lower Colorado River Basin

The Colorado River Basin (CRB) includes seven states and provides municipal and industrial water to millions of people across all major southwestern cities both inside and outside the basin. Agriculture is the largest part of the CRB economy and crop production depends on irrigation, which accounts for about 74% of the total water demand cross the region. A better understanding of irrigation water demands is critically needed as temperatures continue to rise and drought intensifies, potentially leading to water shortages across the region. Yet, past research on irrigation dynamics has generally utilized relatively low spatiotemporal resolution datasets and has often overlooked the relationship between climate and management decisions such as land fallowing, i.e., the practice of leaving cultivated land idle for a growing season. Here, we produced annual estimates of fallow and active cropland extent at high spatial resolution (30 m) from 2001 to 2017 by applying the fallow-land algorithm based on neighborhood and temporal anomalies (FANTA). We specifically focused on diverse CRB agricultural regions: the lower Colorado River planning (LCRP) area and the Pinal and Phoenix active management areas (PPAMA). Utilizing ground observations collected in 2014 and 2017, we found an overall classification accuracy of 88.9% and 87.2% for LCRP and PPAMA, respectively. We then quantified how factors such as climate, district water rights, and market value influenced: (1) annual fallow and active cropland extent and (2) annual cropland productivity, approximated by integrated growing season NDVI (iNDVI). We found that for the LCRP, a region of winter cropping and senior (i.e., preferential) water rights, active cropland productivity was positively correlated with cool-season average vapor pressure deficit (R = 0.72; p &lt; 0.01). By contrast, for the PPAMA, a region of summer cropping and junior water rights, annual fallow and active cropland extent was positively correlated with cool-season aridity (precipitation/potential evapotranspiration) (R = 0.46; p &lt; 0.05), and active cropland productivity was positively correlated with warm-season aridity (precipitation/potential evapotranspiration) (R = 0.42; p &lt; 0.01). We also found that PPAMA cropland productivity was more sensitive to aridity when crop prices were low, potentially due to the influence of market value on management decisions. Our analysis highlights how biophysical (e.g., temperature and precipitation) and socioeconomic (e.g., water rights and crop market value) factors interact to explain seasonal patterns of cropland extent, water use and productivity. These findings indicate that increasing aridity across the region may result in reduced cropland productivity and increased land fallowing for some regions, particularly those with junior water rights.

Quantifying misidentification rates of Europa's microfeatures in low-resolution Galileo imaging

Europa is a compelling target because of its unique geology and potential for habitability. Efforts to map the small-scale global geology of the moon have been hampered by the availability of suitable imaging (≤ 230 m/pix), which covers only 15% of the surface. New data on the abundance of microfeatures (features ≤100 km2 in area) on a global scale and their locations would help put better constraints on the geophysics of the icy world. We mapped microfeatures in low-resolution images of the E15RegMap01 region of Europa and compared our low-resolution dataset to a validated dataset collected using regional mosaic imaging. We evaluated how many features were missed in low-resolution mapping, the feature types to which they belonged, how often misclassifications were made, and how many false-positive features there were in the low-resolution dataset. We found that microchaos, hybrids, and spots were found at relatively high completeness rates when compared to the regional mosaic dataset, but domes and especially pits were almost always missed. Microchaos, hybrids, and spots also had the highest rates of consistent classification between the low-resolution and regional mosaic datasets. Finally, 36% of our total potential feature count from the low-resolution dataset were revealed to be false positives, mostly caused by ridges and shadows that were mistaken for features in low-resolution imaging. These findings quantitatively estimate the likely errors in existing and future global geologic maps that note the presence of microfeatures, offer remedies to minimize these errors, and provide guidance for future Europa mappers.

Deep Learning to Detect Road Distress from Unmanned Aerial System Imagery

Surface distress is an indication of poor or unfavorable pavement performance or signs of impending failure that can be classified into a fracture, distortion, or disintegration. To mitigate the risk of failing roadways, effective methods to detect road distress are needed. Recent studies associated with the detection of road distress using object detection algorithms are encouraging. Although current methodologies are favorable, some of them seem to be inefficient, time-consuming, and costly. For these reasons, the present study presents a methodology based on the mask regions with convolutional neural network model, which is coupled with the new object detection framework Detectron2 to train the model that utilizes roadway imagery acquired from an unmanned aerial system (UAS). For a comprehensive understanding of the performance of the proposed model, different settings are tested in the study. First, the deep learning models are trained based on both high- and low-resolution datasets. Second, three different backbone models are explored. Finally, a set of threshold values are tested. The corresponding experimental results suggest that the proposed methodology and UAS imagery can be used as efficient tools to detect road distress with an average precision score up to 95%.

Human face super-resolution on poor quality surveillance video footage

Most super-resolution (SR) methods proposed to date do not use real ground-truth high-resolution (HR) and low-resolution (LR) image pairs; instead, the vast majority of methods use synthetic LR images generated from the HR images. This approach yields excellent performance on synthetic datasets, but on real-world poor quality surveillance video footage, they suffer from performance degradation. A promising alternative is to apply recent advances in style transfer for unpaired datasets, but state-of-the-art work along these lines has used LR images and HR images from completely different datasets, introducing more variation between the HR and LR domains than necessary. In this paper, we propose methods that overcome both of these shortcomings, applying unpaired style transfer learning methods to face SR but using HR and LR datasets that share important properties. The key is to acquire roughly paired training data from a high-quality main stream and a lower-quality sub-stream of the same IP camera. Based on this principle, we have constructed four datasets comprising more than 400 people, with 1–15 weakly aligned real HR–LR pairs for each subject. We adopt a cycle generative adversarial networks (Cycle GANs) approach that produces impressive super-resolved images for low-quality test images never seen during training. Experiments prove the efficacy of the method. The approach to face SR advocated for in this paper makes possible many real-world applications requiring the extraction of high-quality face images from low-resolution video streams such as those produced by security cameras. Developers of diverse applications such as face recognition, 3D face reconstruction, face alignment, face parsing, human–computer interaction, remote sensing, and access control will benefit from the methods introduced in this work.

Performance evaluation of super resolution algorithms in generating high resolution images using MSE and PSNR

In recent days, requirement of high spatial resolution remote sensing data in various fields has increased tremendously. High resolution satellite remote sensing data is obtained with long focal length optical systems and low altitude. As fabrication of high-resolution optical system and accommodating on the satellite is a challenging task, various alternate methods are being explored to get high resolution imageries. Alternately the high-resolution data can be obtained from super resolution techniques. The super resolution technique uses single or multiple low-resolution mis-registered data sets to generate high resolution data set. Various algorithms are employed in super resolution technique to derive high spatial resolution. In this paper we have compared two methods namely overlapping and interleaving methods and their capability in generating high resolution data are presented.

Video Super-Resolution Based on Generative Adversarial Network and Edge Enhancement

With the help of deep neural networks, video super-resolution (VSR) has made a huge breakthrough. However, these deep learning-based methods are rarely used in specific situations. In addition, training sets may not be suitable because many methods only assume that under ideal circumstances, low-resolution (LR) datasets are downgraded from high-resolution (HR) datasets in a fixed manner. In this paper, we proposed a model based on Generative Adversarial Network (GAN) and edge enhancement to perform super-resolution (SR) reconstruction for LR and blur videos, such as closed-circuit television (CCTV). The adversarial loss allows discriminators to be trained to distinguish between SR frames and ground truth (GT) frames, which is helpful to produce realistic and highly detailed results. The edge enhancement function uses the Laplacian edge module to perform edge enhancement on the intermediate result, which helps further improve the final results. In addition, we add the perceptual loss to the loss function to obtain a higher visual experience. At the same time, we also tried training network on different datasets. A large number of experiments show that our method has advantages in the Vid4 dataset and other LR videos.

Data-driven time series prediction based on multiplicative neuron model artificial neuron network

This paper develops a hybrid approach combining the neural network and the nonlinear filtering to model and predict terrain profiles for both air and ground vehicles. To simplify the neural network structures and reduce the number of synaptic weights and biases, the multiplicative neuron model (MNM) is utilized to describe the relationship between the unknown elevation ahead and the last few height values on the terrain profile. This paper adopts the gradient descent algorithm (GDA) to train the MNM terrain model and stores the MNM parameters into a nonlinear state-space model. The state vector in the state-space model (i.e., parameters of MNM) evolve agilely once absorbing new observations and measurement of elevation values by the Bootstrap Particle Filter (BPF) algorithm. Data-driven predictions on terrain profiles can be achieved through the updated MNM model. This study utilizes two types of terrain profiles to verify the effectiveness of the proposed MNM–BPF approach. Experimental results on two public datasets indicate that the proposed approach not only overcomes the limitations of conventional terrain models that cannot dynamically tune model parameters according to the newly input information, but also provides a simple but effective single-layered network for modeling terrain profiles. The well-trained MNM–BPF model can achieve the lowest root mean square errors (RMSE) (i.e., 17.3211 on the NS profile, 19.0366 on the EW profile) and average error (AE) (i.e., 1.5852 on the NS profile, 0.14885 on the EW profile) in the low-resolution dataset. The lowest RMSE (i.e., 0.16549 on the left profile, 0.29926 on the right profile) and mean absolute error (MAE) (i.e., 0.13467 on the left profile, 0.23933 on the right profile) results are obtained in the high-resolution dataset. Overall, the developed model is superior to the state-of-the-art models in at least four of the six performance metrics and reduces RMSE by 40.8%, 17.2%, 13.1%, and 6.8% on average on the four testing terrain profiles, respectively. The developed approach can be used as a decision tool for the accurate prediction of terrain profiles with different resolutions.

BloodCaps: A capsule network based model for the multiclassification of human peripheral blood cells

Background and Objective: The classification of human peripheral blood cells yields significance in the detection of inflammation, infections and blood cell disorders such as leukemia. Limitations in traditional algorithms for blood cell classification and increased computational processing power have allowed machine learning methods to be utilized for this clinically prevalent task.Methods: In the current work, we present BloodCaps, a capsule based model designed for the accurate multiclassification of a diverse and broad spectrum of blood cells.Results: Implemented on a large-scale dataset of 8 categories of human peripheral blood cells, the proposed architecture achieved an overall accuracy of 99.3%, outperforming convolutional neural networks such as AlexNet(81.5%), VGG16(97.8%), ResNet-18(95.9%) and InceptionV3(98.4%). Furthermore, we devised three new datasets(low-resolution dataset, small dataset, and low-resolution small dataset) from the original dataset, and tested BloodCaps in comparison with AlexNet, VGG16, ResNet-18, and InceptionV3. To further validate the applicability of our proposed model, we tested BloodCaps on additional public datasets such as the All IDB2, BCCD, and Cell Vision datasets. Compared with the reported results, BloodCaps showed the best performance in all three scenarios.Conclusions: The proposed method proved superior in octal classification among all three datasets. We believe the proposed method represents a promising tool to improve the diagnostic performance of clinical blood examinations.

Property Analysis of Adversarially Robust Representation

In this paper, we address the open question: “What do adversarially robust models look at?” Recently, it has been reported in many works that there exists the trade-off between standard accuracy and adversarial robustness. According to prior works, this trade-off is rooted in the fact that adversarially robust and standard accurate models might depend on very different sets of features. However, it has not been well studied what kind of difference actually exists. In this paper, we analyze this difference through various experiments visually and quantitatively. Experimental results show that adversarially robust models look at things at a larger scale than standard models and pay less attention to fine textures. Furthermore, although it has been claimed that adversarially robust features are not compatible with standard accuracy, there is even a positive effect by using them as pre-trained models particularly in low resolution datasets.

Classification of white blood cells using weighted optimized deformable convolutional neural networks

Background Machine learning (ML) algorithms have been widely used in the classification of white blood cells (WBCs). However, the performance of ML algorithms still needs to be addressed for being short of gold standard data sets, and even the implementation of the proposed algorithms. Methods In this study, the method of two-module weighted optimized deformable convolutional neural networks (TWO-DCNN) was proposed for WBC classification. Our algorithm is characterized as two-module transfer learning and deformable convolutional (DC) layers for the betterment of robustness. To validate the performance, our method was compared with classical MLs of VGG16, VGG19, Inception-V3, ResNet-50, support vector machine (SVM), multilayer perceptron (MLP), decision tree (DT) and random forest (RF) on our undisclosed WBC data set and public BCCD data set. Results TWO-DCNN achieved the best performance with the precisions (PREs) of 95.7%, 94.5% and 91.6%, recalls (RECs) of 95.7%, 94.5% and 91.6%, F1-scores (F1s) of 95.7%, 94.5% and 91.6%, area under curves (AUCs) of 0.98, 0.97 and 0.95 for low-resolution and noisy undisclosed data sets, BCCD data set, respectively. Conclusions With accurate feature extraction and optimized network weights, the proposed TWO-DCNN showed the best performance in WBC classification for low-resolution and noisy data sets. It could be used as an alternative method for clinical applications.

Utilizing Advance Texture Features for Rapid Damage Detection of Built Heritage Using High-Resolution Space Borne Data: A Case Study of UNESCO Heritage Site at Bagan, Myanmar

Heritage sites are vulnerable to damage due to social, anthropological and environmental factors. Major Earthquakes are followed by damage to cultural heritage buildings. The assessment of such building damage is a critical problem. Earth Observation data, owing to its property of being non-contact, cost effective, synoptic view and high repeatability, has a significant role to play in estimation of damage in the earthquake affected areas. Currently, several and varied types of remote sensing data have become available, and therefore, appropriate methods for rapid assessment and analysis of the data need to be developed. Rapid damage assessment is critical to minimize loss in terms of life and property. In case of cultural monuments, rapid assessment can minimize damage and help in the conservation of monument. This research focuses on evolving a robust method for rapid identification and extraction of damaged heritage building structures, especially those affected by disasters such as earthquakes. In this study, we propose to examine the utility of advance texture algorithms such as Gabor, fractal and semi-variogram for rapid damage detection in heritage building structures. The methodology attempts to automatically highlight damaged portions of the structure through a knowledge driven rule set. The technique was able to extract the damaged area from the heritage building structure with the use of high-resolution space borne data. It is observed that feature extraction algorithms based on fractal and variogram provide better results than the Gabor based textures and are very useful in the case of high-resolution satellite imagery. Both the methods are able to extract damaged features in both shadowed and non-shadowed regions of the image. Hence the problems posed by shadowed dead grounds on EO data can be effectively resolved. However, it is also observed that the advance texture feature extraction algorithms are useful only in case of high-spatial resolution dataset and has limited use for rapid damage assessment from medium and low resolution datasets.

Biophysical Characterization of Type III Pantothenate Kinase (PanK) from Acinetobacter baumannii

Type-III Pantothenate kinase from the multi drug resistant bacteria, Acinetobacter baumannii (AbPanK) catalyzes the first step of the essential Coenzyme A biosynthesis pathway. AbPanK is an attractive drug target against the bacteria since it is an essential enzyme and its structure is significantly different from the human PanK. AbPanK was cloned, expressed, purified and crystallized. A good quality single crystal was used for X-ray intensity data collection. Dynamic light scattering was done for calculating the hydrodynamic radii and its oligomeric nature in the solution. Binding studies of this protein with its two substrates, Pantothenate and ATP were done using spectrofluorometer. Our results indicated that AbPanK shows a strong affinity with pantothenate with dissociation constant of 1.2 x 10- 8 M and moderate affinity towards ATP of 3.7x 10-3 M. This fact was further substantiated by the calculations of Km of both substrates using kinase assay kit. Dynamic light scattering studies have shown that it exists as homogenous solution with hydrodynamic radii corresponding to the molecular weight of 29.55 kDa. A low-resolution X-ray intensity data set was collected, which shows that AbPank crystallizes in P2 space group with cell dimensions of a= 165 Å, b= 260 Å, and, c= 197 Å and α= 90.0, β= 113.60, γ= 90.0. Recombinant Pantothenate kinase from Acinetobacter baumannii was purified to homogeneity and crystallized. The enzyme exhibits very low sequence identity (28%) to other corresponding enzymes. The recombinant enzyme was active and its binding affinities with its substrates pantothenate and ATP have been studied. This information would be very useful while designing the inhibitors of this enzyme in order to fight bacterial infections associated to this pathogen.

Tree, Shrub, and Grass Classification Using Only RGB Images

In this work, a semantic segmentation-based deep learning method, DeepLabV3+, is applied to classify three vegetation land covers, which are tree, shrub, and grass using only three band color (RGB) images. DeepLabV3+’s detection performance has been studied on low and high resolution datasets that both contain tree, shrub, and grass and some other land cover types. The two datasets are heavily imbalanced where shrub pixels are much fewer than tree and grass pixels. A simple weighting strategy known as median frequency weighting was incorporated into DeepLabV3+ to mitigate the data imbalance issue, which originally used uniform weights. The tree, shrub, grass classification performances are compared when all land cover types are included in the classification and also when classification is limited to the three vegetation classes with both uniform and median frequency weights. Among the three vegetation types, shrub is found to be the most challenging one to classify correctly whereas correct classification accuracy was highest for tree. It is observed that even though the median frequency weighting did not improve the overall accuracy, it resulted in better classification accuracy for the underrepresented classes such as shrub in our case and it also significantly increased the average class accuracy. The classification performance and computation time comparison of DeepLabV3+ with two other pixel-based classification methods on sampled pixels of the three vegetation classes showed that DeepLabV3+ achieves significantly higher accuracy than these methods with a trade-off for longer model training time.

A Temporal Disaggregation Approach for TRMM Monthly Precipitation Products Using AMSR2 Soil Moisture Data

Accurate and spatially-distributed precipitation information is vital to the study of the regional hydrological cycle and water resources, as well as for environmental management. To provide high spatio-temporal resolution precipitation estimates over insufficient rain-gauge areas, great efforts have been taken in using the Normalized Difference Vegetation Index (NDVI) and other land surface variables to improve the spatial resolution of satellite precipitation datasets. However, the strong spatio-temporal heterogeneity of precipitation and the “hysteresis phenomenon” of the relationship between precipitation and vegetation has limited the application of these downscaling methods to high temporal resolutions. To overcome this limitation, a new temporal downscaling method was proposed in this study by introducing daily soil moisture data to explore the relationship between precipitation and the soil moisture increment index. The performance of this proposed temporal downscaling was assessed by downscaling the Tropical Rainfall Measuring Mission (TRMM) precipitation data from a monthly scale to a daily scale over the Hekouzhen to Tongguan of the Yellow River in 2013, and the downscaled daily precipitation datasets were validated with in-situ measurement from 23 rainfall observation stations. The validation results indicate that the downscaled daily precipitation agrees with the rain gauge observations, with a correlation coefficient of 0.59, a mean error range of 1.70 mm, and a root mean square error of 5.93 mm. In general, the monthly precipitation decomposition method proposed in this paper has combined the advantage of both microwave remote sensing products. It has acceptable precision and can generate precipitation on a diurnal scale. It is an important development in the field of using auxiliary data to perform temporal downscaling. Furthermore, this method also provides a reference example for the temporal downscaling of other low temporal resolution datasets.

Attention-aware perceptual enhancement nets for low-resolution image classification

Classifying low-resolution (LR) images is notoriously challenging because of their noisy representation and limited information. Existing approaches mainly solve this challenge by training carefully designed architectures on LR datasets or by employing an image-resizing algorithm in a straightforward manner. However, the performance improvements of these methods are usually limited or even trivial in the case of LR images. In this work, we address the LR image classification problem by developing an end-to-end architecture that internally elevates representations of an LR image to “super-resolved” ones. This approach imparts characteristics similar to those of high-resolution (HR) images and is thus more discriminative and representative for image classification. For this purpose, we propose an innovative unified framework, named Attention-aware Perceptual Enhancement Nets (APEN), which integrates perceptual enhancement and an attention mechanism in an end-to-end manner for LR image classification. Specifically, the framework includes a perceptual enhancement network to generate super-resolved images from LR images. In addition, a novel attention mechanism is presented to highlight informative regions, while restricting the semantic deviation of super-resolved images. Additionally, we designed a feature rectification strategy to promote the adaptability of category decision. Experiments conducted on publicly available datasets demonstrate the superiority of our method against state-of-the-art methods on both LR and HR datasets.

Detecting Cultural Remains in Boreal Forests in Sweden Using Airborne Laser Scanning Data of Different Resolutions

ABSTRACTAirborne laser scanning (ALS) is increasingly being used to detect cultural remains in forest landscapes. Boreal forests are challenging, however, since most ancient land use was carried out without major permanent ground disturbances. If this challenge can be met, there is a large potential for surveys through existing nation-wide laser-scanning programs, although their resolutions tend to be low. In this study, we compare the performance of low-resolution (LR) and high-resolution (HR) ALS data in the Krycklan catchment in northern Sweden, an area where ancient land use was small-scale and diverse. About three times as many cultural remains were detected in the HR data set, but the LR set was satisfactory for distinct structures. We analyze how LR data sets can be enhanced at ground-point classification and terrain-model generation and conclude that ALS data have a large potential for the detection and protection of cultural remains in the boreal forest.