Water Areas Segmentation from Remote Sensing Images Using a Separable Residual SegNet Network

Compared with natural images, remote sensing targets have small and dense target shapes as well as complex target backgrounds. As a result, insufficient detection accuracy and target location cannot be accurately identified. So, this paper proposes the YOLO-extract algorithm based on the YOLOv5 algorithm. Firstly, The YOLO-extract algorithm optimized the model structure of the YOLOv5 algorithm. The YOLO-extract algorithm not only deleted the feature layer and prediction head with poor feature extraction ability but also a new feature extractor with stronger feature extraction ability was integrated into the network. At the same time, YOLO-extract borrowed the idea of residual network to integrate Coordinate Attention into the network. Secondly, the mixed dilated convolution was combined with the redesigned residual structure to enhance the feature and location information extraction ability of the shallow layer of the model and optimize the feature extraction ability of the model for different scale targets. Finally, drawing on the idea of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula> -IoU Loss, Focal- <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula> EIoU Loss was designed to replace CIoU Loss, which makes the model bounding box regression faster and the loss lower. The experimental results on the test data set show that compared with the YOLOv5 algorithm, the YOLO-extract algorithm has a faster convergence speed, reduces the calculation amount by 45.3GFLOPs and the number of parameters by 10.526M, but increases the mAP by 8.1% and the detection speed by 3 times.

With the wide applications of remote sensing technology in engineering, the demands of efficient object detection algorithms for remote sensing images have also been significantly increased in recent years. Traditional detection methods have the shortcomings of low accuracy and poor robustness, which is difficult to be applied to remote sensing images with complex background and varieties of objects. Recently, the deep convolutional neural networks have already shown great advances in object detection and outperformed many traditional methods. In this work, we study the performance of a region proposal-based method, Mask R-CNN, for detecting airplane and ship in remote sensing images. Specifically, we add an FPN module to improve the accuracy of small objects detection, and a mask branch is used to describe the shape of objects. In addition, we used a series of data augmentation strategies during training for meeting the CNN’s requirement of training samples diversity. The experimental results show that our model has superior performance in object detection of remote sensing images.

Modelling species distributions with remote sensing data: bridging disciplinary perspectives

Although the existing traditional image classification methods have been widely applied in practical problems, there are some problems in the application process, such as unsatisfactory effects, low classification accuracy, and weak adaptive ability. This method separates image feature extraction and classification into two steps for classification operation. The deep learning model has a powerful learning ability, which integrates the feature extraction and classification process into a whole to complete the image classification test, which can effectively improve the image classification accuracy. However, this method has the following problems in the application process: first, it is impossible to effectively approximate the complex functions in the deep learning model. Second, the deep learning model comes with a low classifier with low accuracy. So, this paper introduces the idea of sparse representation into the architecture of the deep learning network and comprehensively utilizes the sparse representation of well multidimensional data linear decomposition ability and the deep structural advantages of multilayer nonlinear mapping to complete the complex function approximation in the deep learning model. And a sparse representation classification method based on the optimized kernel function is proposed to replace the classifier in the deep learning model, thereby improving the image classification effect. Therefore, this paper proposes an image classification algorithm based on the stacked sparse coding depth learning model-optimized kernel function nonnegative sparse representation. The experimental results show that the proposed method not only has a higher average accuracy than other mainstream methods but also can be good adapted to various image databases. Compared with other deep learning methods, it can better solve the problems of complex function approximation and poor classifier effect, thus further improving image classification accuracy.

In land-use classification of hyperspectral remote sensing (RS) images, traditional classification methods often experience large amount of datasets and low efficiency. To solve these problems, a fast machine-learning method, the extreme learning machine (ELM) algorithm, was introduced. However, basic use of the ELM usually encounters problems of unstable classification results and low classification accuracy. Hence, in this paper, optimization methods for ELM-based RS image classification were mainly discussed and applied to solve the bottleneck problems. From the three perspectives of ensemble learning, making full use of image texture features, and deep learning, three classification optimization methods have been designed and implemented. The results show that: 1) To some extent, all the three methods can achieve a balance between classification accuracy and efficiency, i.e., they can maintain the advantage of ELM algorithm in classification efficiency and speed while have better classification accuracy; 2) The image texture feature optimization method (LBP-KELM) solves the problem of unsatisfactory classification results experienced by the ensemble learning optimization method (Ensemble-ELM) and further improves classification accuracy. However, the classification results are sensitive to the type of dataset; and 3) Fortunately, the optimization method combined with deep learning (CNN-ELM) can meet the application needs of multiple datasets. Furthermore, it can also further improve classification accuracy.

In view of the advantages of deep learning model in the extraction of abstract concept, a new text clustering algorithm is designed based on Deep Boltzmann Machines. Based on Replicate Softmax Model and new Deep Boltzmann Machine, energy function of this model is proposed and the detail learning algorithm is introduced.

Traditional engine fault diagnosis methods usually need to extract the features manually before classifying them by the pattern recognition method, which makes it difficult to solve the end-to-end fault diagnosis problem. In recent years, deep learning has been applied in different fields, bringing considerable convenience to technological change, and its application in the automotive field also has many applications, such as image recognition, language processing, and assisted driving. In this paper, a one-dimensional convolutional neural network (1D-CNN) in deep learning is used to process vibration signals to achieve fault diagnosis and classification. By collecting the vibration signal data of different engine working conditions, the collected data are organized into several sets of data in a working cycle, which are divided into a training sample set and a test sample set. Then, a one-dimensional convolutional neural network model is built in Python to allow the feature filter (convolution kernel) to learn the data from the training set and these convolution checks process the input data of the test set. Convolution and pooling extract features to output to a new space, which is characterized by learning features directly from the original vibration signals and completing fault diagnosis. The experimental results show that the pattern recognition method based on a one-dimensional convolutional neural network can be effectively applied to engine fault diagnosis and has higher diagnostic accuracy than traditional methods.

Remote sensing images are blurred due to large and wide imaging, long shooting distance, fast scanning speed, interference from external light, etc. At the same time, because of that remote sensing images have the characteristics of diverse and dense shooting objects, deblurring remote sensing images is a major problem in remote sensing research. Therefore, we propose a remote sensing image deblurring algorithm, which based on WGAN. The algorithm is different from the traditional method in estimating the blur kernel of image. What’s more our method does not require an explicit estimation of the blur kernel, and it implements an end-to-end image deblurring process. We use a WGAN-based deblurring model. First, the training images are processed in pairs. Then, in order to increase the generalization ability, a image of 256 * 256 that is a sub-region cropped at the random position in the original image is chosen as the input image. Finally, to achieve a better deblurring effect, a content loss function and a perceptual loss function are added to the loss function to achieve the specific implementation. The remote sensing image deblurring model trained by the proposed method has achieved better results on the remote sensing image dataset. The experimental results show that the proposed algorithm have better performance than the traditional method in filtering out the blur of remote sensing images, which could optimize the overall visual effect subjectively and improve the peak signal-to-noise ratio of the image objectively.

Aiming at the problem of low target detection accuracy caused by the close arrangement of targets, complex feature information, and numerous small targets in remote sensing images, this paper proposes a target detection algorithm based on the improved Yolov5 algorithm for remote sensing images. Firstly, this paper designs a remote sensing image target detection algorithm C²-YOLO based on multi-level feature fusion. Combined with the attention mechanism module to focus on the spatial location information of the target, which improves the extraction effect of the features of the target of interest. The network uses the CARAFE upsampling operator with a larger sensing field to better utilize the surrounding information. The proposal is to incorporate a predictive header hl that is generated from a low-level, high-resolution feature map to improve the recognition of tiny objects. The problem of large target scales in remote sensing images and poor detection of small targets has been improved. Angular loss is also added to the loss function to improve the detection of rotating targets. Finally, the accuracy comparison and experimental effect visualization are carried out on the dataset, and the experimental results show that the improvement work in this paper achieves better results no matter comparing the original YOLOv5 detection algorithm or other types of detection algorithms. Specifically, the accuracy of C²-YOLO on the NWPU dataset 1-shot recognition task is 78.45%, which is much higher than that of other detection algorithms, and is able to reach a superior performance in the remote sensing image target detection task.

We propose a novel machine learning model for classification problems, Deep Twin Support Vector Machine (DTWSVM), which combines TWSVM with deep learning ideas. TWSVM is a successful algorithm for classification problems which seeks two nonparallel hyper planes to make each hyper plane close to one class and far from the other as much as possible. And Deep Learning (DL) models have shown good ability in feature extraction and dimension reduction by constructing multi-layer network. Since the feature extraction in DL can reduce feature dimension while maintain the main information of original inputs, we consider constructing a three layer network which contains input layer, hidden layer and output layer. We use two TWSVMs in the hidden layer to extract features based on the projection principle which is derived from Multi-Layer Perceptron (MLP). The two TWSVMs will get four hyper planes by solving four convex quadratic programs. A new dataset which consists of the extracted features with four feature dimensions is produced from the hidden layer and then we can input it to the main TWSVM of the output layer to make final prediction. Similar as TWSVM, we design linear DTWSVM and nonlinear DTWSVM which have been proved to be very effective in classification problems. In numerical experiments, we have obtained 100% prediction accuracy for several datasets which is state-of-the-art performance absolutely!

Abstract. Utilizing ground control points (GCPs) to georeference photogrammetry-based point cloud data is a common practice in unmanned aerial system (UAS) mapping. Direct georeferencing or integrated sensor orientation (ISO) can be used to obtain georeferenced point clouds from UAS without relying heavily on GCPs. However, the accuracy of the point cloud may be impacted by the accuracy of the trajectory solution obtained by GNSS. To improve point cloud accuracy, post-processing kinematic (PPK) solutions can be applied to the UAS trajectory, which may provide higher accuracy than low-accuracy trajectory solutions and minimize the reliance on GCPs. This study compares the accuracy and precision of two different point clouds generated using different methods. One point cloud was generated using traditional photogrammetric methods with low accuracy Global Navigation Satellite System (GNSS) observations from the UAS and GCPs that have an average accuracy of one to two centimeters, while the other was generated using PPK trajectory solution for the UAS’s trajectory with two software: open-source Emlid Studio and the widely used Inertial Explorer. The use of PPK techniques in UAS mapping may have several potential benefits over traditional methods. By correcting the errors in the UAS's trajectory, a user may only need to depend on fewer ground control points, which can reduce the time and cost associated with fieldwork. This is particularly useful in areas that are difficult to access or have limited ground control point options, such as in urban or forested areas. To evaluate performance, a GNSS receiver is used to obtain measurements on checkpoints, which are used to assess the accuracies of the point clouds. In our experiments, the accuracy of the point clouds generated using PPK trajectory solution with high accuracy GCPs was found to be higher than those generated with low accuracy GNSS observations while aided with high accuracy ground control points. While the use of PPK with GCPs is generally expected to provide more accurate and reliable data than low-accuracy GNSS observations even after adjusting with GCPs, the number and distribution of GCPs can still significantly impact overall accuracy. Therefore, careful consideration of the number of GCPs and their placement is essential to achieve the desired level of efficiency and effectiveness in UAS mapping.

Bug triage is the process of quickly assigning bug reports to the most suitable developer so that the defect can be fixed quickly. Usually, the manual triage method not only consumes a lot of time for developers, but also has low ac-curacy. Therefore, how to automatically and efficiently allocate defects has become a hot research problem in the field of software maintenance. At present, the automatic bug triage method has become the mainstream technology, but there are also problems such as feature extraction and insufficient text representation ability. In view of the above problems, the task of automatic bug triage is regarded as a text classification problem. The method for automatic bug triage of multi-scale convolutional neural networks based on word vectors MCNN- BT can more effectively capture the important text features of bug reports, thereby providing more accurate defect repairer recommendation services. Empirical research is carried out in five software projects, Platform, Pde, Jdt, Cdt, Birt. The experimental results show that the prediction accuracy of MCNN-BT method is higher than that of the Naive Bayes method and the Latent Dirichlet Allocation(LDA) method.

AN ASSESSMENT OF SPATIAL VARIATION OF LAND SURFACE CHARACTERISTICS OF MINNA, NIGER STATE NIGERIA FOR SUSTAINABLE URBANIZATION USING GEOSPATIAL TECHNIQUES

According to 230 flux site observations, intra-day and diurnal ground heat flux (G) accounted for 19.2 % and 28.8 % of the corresponding net radiation, respectively. This indicates that G plays an important role in remote sensing (RS) evapotranspiration (ET) models. The accuracy of the five G simulation methods in the surface energy balance-based RS ET models was evaluated using half-hourly observations. The linear coefficient (LC) method and the methods embedded with the normalized difference vegetation index (NDVI) were able to accurately simulate a half-hourly G series at most sites. The mean and median Nash-Sutcliffe efficiency (NSE) values of all sites were generally higher than 0.50 in each half-hour period. However, the two methods embedded by fractional vegetation coverage showed poor performance in most half-hour periods, except the sunrise and sunset periods, with the mean and median NSE values of all sites below 0.20. The accuracy of each method varied significantly at different sites and at half-hour intervals. The highest accuracy was exhibited during sunrise periods (6:00–7:00), followed by sunset periods (17:00–18:00). There were 92 % (211/230) sites with an NSE of the LC method greater than 0.50 at 6:30. It showed a slightly higher accuracy during night periods than during daytime periods. The lowest accuracy was observed at noon periods (10:00–15:30). The sites with an NSE exceeding 0.50 only accounted for 51 % (118/230) and 43 % (100/230) at 10:30 and 13:30, respectively. The accuracy of Northern Hemisphere sites was generally higher than that of Southern Hemisphere sites. In general, the highest and lowest accuracies were observed at the high- and low-latitude sites, respectively. The performance of the LC method and the methods embedded with NDVI were generally satisfactory at the Eurasian and North American sites, with the NSE values of most sites exceeding 0.70. Conversely, it exhibited relatively poor performance at the African, South American, and Oceanian sites, especially the African sites. Both the temporal and spatial distributions of the accuracy of the G simulation were positively correlated with the correlation between G and the net radiation. Although the G simulation methods accurately simulated the G series at most sites and time periods, their performance was poor at some sites and time periods. The application of RS ET datasets covering these sites requires caution. The optimal parameter value for each method varied greatly at different sites. This indicates that the fixed parameter values in the G simulation methods are not appropriate. Further improvement of G simulations at these sites and time periods is recommended for the RS ET modelers.

Semantic segmentation in remote sensing images is a very challenging task. Each pixel in a remote sensing image has a semantic meaning to it and automatic annotation of each pixel remains as an open challenge for the research community due to its high spatial resolution. To address this issue deep learning based encoder-decoder architectures like SegNet and ResNet that is widely used for computer vision dataset is adopted for remote sensing images and its performance is analyzed based on the pixel wise classification accuracy. From the experiment conducted it is inferred that SegNet suffers from degradation problem when the depth of the network is increased with an overall accuracy of about 86.086% whereas the Residual network manages to overcome the degradation effect with an overall accuracy of about 87.747%.