Figure - available from: Multimedia Tools and Applications
This content is subject to copyright. Terms and conditions apply.
Source publication
Land-use classification using remote sensing images plays a key role in many applications such as urban mapping and geospatial object detection. With the rapid development of satellite sensors, high-resolution images which exhibit more detailed textures now can be acquired. How to effectively represent these images and recognize the categories of l...
Similar publications
In this paper, we propose a novel framework for object detection and recognition in cluttered images, given a single hand-drawn example as model. Compared with previous work, our contribution is three-fold. 1) Three preprocessing procedures are proposed to reduce the number of irrelevant edge fragments that are often generated during edge detection...
![Figure 1. First flight control models [3]](profile/Jihad-Anwar/publication/342946036/figure/fig16/AS:913468576264192@1594799288040/First-flight-control-models-3_Q320.jpg)
Humans detection and tracking in videos are difficult and challenging in nature. It is an
essential task to many applications, such as video surveillance and security, intelligent
vehicles, etc.
In this article, Histogram of Oriented Gradient (HOG) algorithm is proposed to find robust set
of features that allows object form to be discriminated. It...
With the rapid increase of optical remote sensing (RS) images, on-orbit specific target detection has been facing more and more challenges: limited processing environment, wireless bandwidth, space, power and storage. To resolve this issue, visual characteristics in the compressed domain are exploited to represent and recognise specific objects. In...
Feature extraction involves feature detection, description and matching which is the baseline of many computer vision applications like content based image retrieval, image classification, image recognition, object detection etc. Features detected should have greater repeatability and should be able to derive descriptors out of it that are highly d...
Citations
... Remote Sensing provides perennial images as a useful source for Land-use/Land-cove Time-Series (LU/LC-TSC) as a result of recent developments in Earth Observation (EO) technology. Unlabeled time-series images of remote sensing are used by terrestrial remote sensing to labeling for showing and predicting Landuse/Land-cover (LU/LC) changes (Chen et al. 2017;Gómez et al. 2016;Tian et al. 2017). Remotely sensed LULC map derives information of LULC changes on the ground (change detection) by assigning each pixel of the image to real-word LULC feature types (Paul et al. 2018). ...
Observing the state and changes of land cover (LC) is critical for assessing the status and changes of ecosystems such as urban development. Remote sensing can extract useful data from a large region regularly and at the same time. The present study has used Landsat 5 and 8 images to derive LC classification and Land Surface Temperature (LST) maps of Reykjavik city for four years 1987, 1995, 2006, and 2019. Satellite images were classified into four major land-cover classes using the Artificial Neural Network classification (ANN) algorithm. Then, calculations including the LST and the distance from the water were calculated for each study period. Finally, using the Markov chain, the prediction of land cover classes and LST classes for 2020 was obtained. The results have shown an average of 21.5% (8.2 km2), 71% (32.9 km2), -62.9% (43.7 km2), and 3.4% (2.6 km2) growth for Urban Areas (UA), Vegetation Land (VL), Bare Lands (BL), and Water Lands (WL) of Reykjavik evaluated LC classes. Reykjavik LC forecasting shows that UA will have had a 38.5% (8.2 km2), VL a -43% (34.1 km2), BL a 14.7% (3.8 km2), and WL a 14.8% (11.9 km2) by 2030. The mean of LST in this area has risen by 67.5% (7.7 °C) during this period. This is also supported by the findings of the analysis of the NDVI and LST connection. According to the research, the LST and the NDVI value of 0.1 to 0.7 have a positive relationship in this area. However, there is a negative correlation between LST with NDVI values of 0.8 and 0.9. The results showed that the mean of LST was also influenced by the distance from WL water bodies. The LST has grown by a mean of 0.7 of correlation as the distance from WL has increased. The results of this research will be useful in order to get an overview of the past and present situation and plan for the control of heat islands and land cover changes in Reykjavik.
... Land-use classification is progressed either using supervised or unsupervised approaches [10,11]. Clustering methods are a form of unsupervised classification that undergoes unclear grouping in the absence of sample sets. ...
... The multi-class SVM employed for the land use multi-classification was more effective than the traditional method and was capable of dealing with the multi-textured images, but suffered from noises. Tian et al. [11] developed Sparse coding Spatial Pyramid Matching (ScSPM) for classifying the land use, for which the features were obtained using Scale Invariant Feature Transform (SIFT), and the features were classified using linear kernel SVM. The merit of the method was that the method was more precise and concise, but the method failed because of high computational complexity. ...
Remote sensing plays a major role in crop classification, land use classification, and land cover classification such that the information for the classification is assured with the help of the satellite images. This paper concentrates on the land use classification and proposes an optimization algorithm, called Firefly Harmony Search (FHS) for training the Deep Belief Neural Network (DBN). The FHS algorithm is the integration of the Firefly Algorithm and Harmony Search Algorithm (HSA), which tunes the weights of DBN to perform the multi-class classification. For the effective classification, the multispectral image is subjected to the sparse Fuzzy C-Means to form segments such that the feature extraction is effective, free from dimensionality issues and computational complexities. The features extracted from the segments of the multi-spectral images include vegetation indices and statistical features. Then, these features are fed to the DBN, which is tuned using the FHS algorithm for performing the land use classification. Experimentation using four datasets proves the effectiveness of the proposed multi-class classification approach. The accuracy, sensitivity, and specificity of the method are found to be 0.9317, 0.9568, and 0.0379, respectively, that is effective over the existing land use classification methods.
... A new scene classification method has been proposed in [4] that combines the theorybased multiple CNNs models to achieve better accuracy than older methods. The research in [5] proposed a sparse coding method for land use classification. This method is biologically motivated and novel in its combination with spatial pyramid matching (ScSPM). ...
Developing remote sensing technology enables the production of very high-resolution (VHR) images. Classification of the VHR imagery scene has become a challenging problem. In this paper, we propose a model for VHR scene classification. First, convolutional neural networks (CNNs) with pre-trained weights are used as a deep feature extractor to extract the global and local CNNs features from the original VHR images. Second, the spectral residual-based saliency detection algorithm is used to extract the saliency map. Then, saliency features from the saliency map are extracted using CNNs in order to extract robust features for the VHR imagery, especially for the image with salience object. Third, we use the feature fusion technique rather than the raw deep features to represent the final shape of the VHR image scenes. In feature fusion, discriminant correlation analysis (DCA) is used to fuse both the global and local CNNs features and saliency features. DCA is a more suitable and cost-effective fusion method than the traditional fusion techniques. Finally, we propose an enhanced multilayer perceptron to classify the image. Experiments are performed on four widely used datasets: UC-Merced, WHU-RS, Aerial Image, and NWPU-RESISC45. Results confirm that the proposed model performs better than state-of-the-art scene classification models.
... A new scene classification method has been proposed in [4] that combines the theorybased multiple CNNs models to achieve better accuracy than older methods. The research in [5] proposed a sparse coding method for land use classification. This method is biologically motivated and novel in its combination with spatial pyramid matching (ScSPM). ...
Developing remote sensing technology enables the production of very high-resolution (VHR) images. Classification of the VHR imagery scene has become a challenging problem. In this paper, we propose a model for VHR scene classification. First, convolutional neural networks (CNNs) with pre-trained weights are used as a deep feature extractor to extract the global and local CNNs features from the original VHR images. Second, the spectral residual-based saliency detection algorithm is used to extract the saliency map. Then, saliency features from the saliency map are extracted using CNNs in order to extract robust features for the VHR imagery, especially for the image with salience object. Third, we use the feature fusion technique rather than the raw deep features to represent the final shape of the VHR image scenes. In feature fusion, discriminant correlation analysis (DCA) is used to fuse both the global and local CNNs features and saliency features. DCA is a more suitable and cost-effective fusion method than the traditional fusion techniques. Finally, we propose an enhanced multilayer perceptron to classify the image. Experiments are performed on four widely used datasets: UC-Merced, WHU-RS, Aerial Image, and NWPU-RESISC45. Results confirm that the proposed model performs better than state-of-the-art scene classification models.
... Land-Use/Land-Cover Time-Series Classification (LULC-TSC), which is an important and challenging problem in terrestrial remote sensing, uses multiple labeled time-series images for training to predict LULC class labels of unlabeled time-series remote sensing images (Gomez et al., 2016;Chen et al., 2017;Tian et al., 2017). Particularly since the development of Earth Observation (EO) technology, perennial accumulated remote-sensing images have been providing usable data sources for LULC-TSC. ...
Land-Use/Land-Cover Time-Series Classifcation (LULC-TSC) is an important and challenging problem in terrestrial remote sensing. Detecting change-points, dividing the entire time series into multiple invariant subsequences, and classifying the subsequences can improve LULC classifcation efciency. Therefore, we have
proposed a Time-Series Classifcation approach based on Change Detection (TSCCD) for rapid LULC mapping
that uses the Prophet algorithm to detect the ground-cover change-points and perform time-series segmentation
in a time dimension and the DTW (Dynamic Time Warping) algorithm to classify the sub-time series. Since we
can assume that the ground cover remains unchanged in each subsequence, only one time-training sample
selection and one LULC classifcation are needed, which greatly improves the work efciency. Prophet can
accurately detect large and subtle changes, capture change direction and change rate, and is strongly robust for
handling noise and missing data. DTW is mainly used to improve the accuracy of time-series classifcation and to
resolve the time misalignment problems of ground-cover series data caused by irregular observations or missing
values. The results of comparative experiments with BFAST, LandTrendR, and CCDC using simulated time-series
showed that TSCCD can detect large and subtle changes and capture change direction and change rate, performing substantially better than the other three contrasting algorithms overall in time-series change detection.
Finally, the MODIS (Moderate Resolution Imaging Spectroradiometer) time-series images of Wuhan City from
2000 to 2018 were selected for TSCCD, and the results of China’s national land-use surveys in 2000, 2005, 2008,
2010, 2013, and 2015 were used for cross-validation. The results showed that the classifcation accuracy of each
tested subsequence was higher than 90% and that most Kappa coefcients were greater than 0.9. This means that
the proposed TSCCD approach can effectively solve real LULC-TSC problems and has high application value. It
can be used for large-area, long time-series LULC classifcation, which is of great guiding signifcance for
studying global environmental changes, forest-cover changes, and conducting land-use surveys.
... The authors in [28] combined active learning with PCA, LDA, KPCA and NPE (i.e. four manifold learning methods) to classify remote sensing multi/hyper spectral data [34] and to obtain effective classification accuracy. In [15], a multi-manifold structure for unsupervised hyper-spectral clustering analysis is designed to achieve better accuracy compared with conventional dimensionality reduction techniques. ...
In our increasingly “data-abundant” society, remote sensing big data perform massive, high dimension and heterogeneity features, which could result in “dimension disaster” to various extent. It is worth mentioning that the past two decades have witnessed a number of dimensional reductions to weak the spatiotemporal redundancy and simplify the calculation in remote sensing information extraction, such as the linear learning methods or the manifold learning methods. However, the “crowding” and mixing when reducing dimensions of remote sensing categories could degrade the performance of existing techniques. Then in this paper, by analyzing probability distribution of pairwise distances among remote sensing datapoints, we use the 2-mixed Gaussian model(GMM) to improve the effectiveness of the theory of t-Distributed Stochastic Neighbor Embedding (t-SNE). A basic reducing dimensional model is given to test our proposed methods. The experiments show that the new probability distribution capable retains the local structure and significantly reveals differences between categories in a global structure.
Purpose
Gastrointestinal endoscopy has become a widely used technique to diagnose intestinal anomalies like polyp. However, the vast quantities of images produced in the detection process greatly increase the burden of clinicians and the misdiagnosis rate. To tackle this problem, a new computer-assisted detection system called Salient Codebook Locality-Constrained Linear Coding (SCLLC) with Annular Spatial Pyramid Matching (ASPM) is proposed in this paper to classify intestinal polyp images automatically.
Methods
Firstly, SIFT features are extracted from images and k-means clustering method is utilized on the patch features to obtain the initial codebook. Secondly, the proposed SCLLC algorithm is employed to achieve the salient codebook and encode the features by emphasizing the salient basis vectors in feature coding process. Then, a max-pooling strategy of annular region segmentation based on Spatial Pyramid Matching is proposed to improve the effectiveness of processing for intestinal images. Finally, SVM classifier is developed to carry out polyp images classification tasks.
Results
The experimental results exhibit promising 94.10% accuracy, 91.20% sensitivity and 97.01% specificity. In addition, the 0.97 s detecting time indicates the feasibility of clinical application.
Conclusion
Our proposed method can effectively improve the overall performance of intestinal polyp recognition by integrating the internal base vector correlation of codebook and considering the annular structure of the intestinal images.
Cette thèse aborde la question du traitement correct et complet de la couleur selon les contraintes métrologiques. Le manque d’approches adaptées a justifié la reformulation principaux outils de traitement d’images que sont le gradient, la détection et la description de points d’intérêt. Les approches proposées sont génériques : indépendantes du nombre de canaux d’acquisition (de la couleur à l’hyper-spectral), de la plage spectrale considérée et prenant en compte les courbes de sensibilité spectrales du capteur ou de l’œil.Le full-vector gradient nait de cet objectif métrologique. La preuve de concept est effectuée sur des images couleurs, multi et hyper-spectrales. L’extension développée pour l’analyse de la déficience visuelle ouvre également de nombreuses s perspectives intéressantes pour l’analyse du système visuel humain. Ce gradient est au cœur de la proposition d’un détecteur de points d’intérêt, lui aussi générique. Nous montrons la nécessité d’un choix mathématiquement valide de la distance entre attributs et l’importance de la cohérence de la paire attribut/distance. Une paire attribut/distance complète l’ensemble.Pour chaque développement, nous proposons des protocoles objectifs de validation liés à des générateurs d’images de synthèse explorant toute la complexité spatio-chromatique possible. Notre hypothèse est que la difficulté d’extraction du gradient/des points d’intérêts… est liée à la complexité de discrimination des distributions couleur dans la zone de traitement. Une confrontation aux approches courantes du domaine a été également mise en œuvre.
As a typical representative of big data, geographic spatiotemporal big data present new features especially the non-stationary feature, bringing new challenges to mine correlation information. However, representation of instantaneous information is the main bottleneck for non-stationary data, but the traditional non-stationary analysis methods are limited by Heisenberg's uncertainty principle. Therefore, we firstly represent instantaneous frequency of geographic spatiotemporal big data based on Hilbert-Huang transform to overcome traditional methods' weakness. Secondly, we propose absolute entropy correlation analysis method based on KL divergence. Finally, we select five geographic factors to certify that the absolute entropy correlation analysis method is effective and distinguishable.
