Large-scale Remote Sensing Image Research Articles

Step-by-Step: Efficient Ship Detection in Large-Scale Remote Sensing Images

Step-by-Step: Efficient Ship Detection in Large-Scale Remote Sensing Images

Fully Connected Hashing Neural Networks for Indexing Large-Scale Remote Sensing Images

With the emergence of big data, the efficiency of data querying and data storage has become a critical bottleneck in the remote sensing community. In this letter, we explore hash learning for the indexing of large-scale remote sensing images (RSIs) with a supervised pairwise neural network with the aim of improving RSI retrieval performance with a few binary bits. First, a fully connected hashing neural network (FCHNN) is proposed in order to map RSI features into binary (feature-to-binary) codes. Compared with pixel-to-binary frameworks, such as DPSH (deep pairwise-supervised hashing), FCHNN only contains three fully connected layers and incorporates another new constraint, so it can be significantly accelerated to obtain desirable performance. Second, five types of image features, including mid-level and deep features, were investigated in the learning of the FCHNN to achieve state-of-the-art performances. The mid-level features were based on Fisher encoding with affine-invariant local descriptors, and the deep features were extracted by pretrained or fine-tuned CNNs (e.g., CaffeNet and VGG-VD16). Experiments on five recently released large-scale RSI datasets (i.e., AID, NWPU45, PatternNet, RSI-CB128, and RSI-CB256) demonstrated the effectiveness of the proposed method in comparison with existing handcrafted or deep-based hashing methods.

Object Detection in Large-Scale Remote Sensing Images With a Distributed Deep Learning Framework

With the accumulation and storage of remote sensing images in various satellite data centers, the rapid detection of objects of interest from large-scale remote sensing images is a current research focus and application requirement. Although some cutting-edge object detection algorithms in remote sensing images perform well in terms of accuracy (mAP), their inference speed is slow and requires high hardware requirements that are not suitable for real-time object detection in large-scale remote sensing images. To address this issue, we propose a fast inference framework for object detection in large-scale remote sensing images. On the one hand, we introduce <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 on the YWCSL model to implement adaptive weighted loss and gradient, which achieves 64.62% and 79.54% mAP on DIOR-R and DOTA test sets, respectively. More importantly, the inference speed of the YWCSL model reaches 60.74 FPS on a single NVIDIA GeForce RTX 3080 Ti, which is 2.87 times faster than the current state-of-the-art one-stage detector S <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$^{2}$</tex-math></inline-formula> A-Net. On the other hand, we build a distributed inference framework to enable fast inference on large-scale remote sensing images. Specifically, we save the images on HDFS for distributed storage and deploy the pre-trained YWCSL model on the Spark cluster. In addition, we use a custom partitioner RankPartition to repartition the data to further improve the performance of the cluster. When using 5 nodes, the speedup of the cluster reaches 9.54, which is 90.80% higher than the theoretical linear speedup (5.00). Our distributed inference framework for large-scale remote sensing images significantly reduces the dependence of object detection on expensive hardware resources, which has important research significance for the wide application of object detection in remote sensing images.

An Anchor-free Object Detection Network for Arbitrarily-orientated Ships in Large-scale Remote Sensing Images

ABSTRACT Reliable and efficient ship detection is still a challenging task for large-scale remote sensing images with complex background. Since ships in remote sensing images are usually densely distributed with different sizes and uncertain directions, adopting the common horizontal bounding box to detect the inclined ships with large aspect ratio will lead to inaccurate detection. In this letter, a multitask learning object detection framework called MSRDNet is proposed to detect ships with arbitrarily orientations. Convolutional features are extracted by the backbone with parallel attention residual block (PARB), which combines spatial and channel attention module. Then, a lightweight classifier is designed as a subtask to filter out a large number of background patches. Moreover, the anchor-free detector with orientation regression is designed as the other subtask for the inclined and densely distributed ships. The proposed algorithm and comparative methods are tested on two public remote sensing image data sets DOTA and HRSC2016. The experimental results demonstrate that the proposed algorithm outperforms all the compared methods. The code is available from https://github.com/Sdy344/paper_code .

Object classification of remote sensing images based on optimized projection supervised discrete hashing

Recently, with the number of large-scale remote sensing (RS) images increasing, the demand for large-scale RS image object classification is growing, and many researchers are interested. Hashing, as a result of its low memory requirements and high time efficiency, has widely solved the problem of large-scale RS images. Supervised hashing methods mainly leverage RS image label information to learn hashing function; however, the similarity of the original feature space cannot be well preserved, which cannot meet the accurate requirements of RS images object classification. To address the aforementioned analysis, we propose a method named optimized projection supervised discrete hashing (OPSDH), which jointly learns optimized projection constraint and discrete binary codes generation model. It uses an effective optimized projection method to further constraint on the supervised hashing learn, and generated hash codes preserve the similarity based on the data label while retaining the original feature space’s similarity. The experimental results show that OPSDH reaches improved performance compared with existing hashing methods and demonstrates that the proposed OPSDH is more efficient for operational applications.

Simultaneous Road Surface and Centerline Extraction From Large-Scale Remote Sensing Images Using CNN-Based Segmentation and Tracing

Accurate and up-to-date road maps are of great importance in a wide range of applications. Unfortunately, automatic road extraction from high-resolution remote sensing images remains challenging due to the occlusion of trees and buildings, discriminability of roads, and complex backgrounds. To address these problems, especially road connectivity and completeness, in this article, we introduce a novel deep learning-based multistage framework to accurately extract the road surface and road centerline simultaneously. Our framework consists of three steps: boosting segmentation, multiple starting points tracing, and fusion. The initial road surface segmentation is achieved with a fully convolutional network (FCN), after which another lighter FCN is applied several times to boost the accuracy and connectivity of the initial segmentation. In the multiple starting points tracing step, the starting points are automatically generated by extracting the road intersections of the segmentation results, which then are utilized to track consecutive and complete road networks through an iterative search strategy embedded in a convolutional neural network (CNN). The fusion step aggregates the semantic and topological information of road networks by combining the segmentation and tracing results to produce the final and refined road segmentation and centerline maps. We evaluated our method utilizing three data sets covering various road situations in more than 40 cities around the world. The results demonstrate the superior performance of our proposed framework. Specifically, our method’s performance exceeded the other methods by 7% and 40% for the connectivity indicator for road surface segmentation and for the completeness indicator for centerline extraction, respectively.

Research on Improved Method of Storage and Query of Large-Scale Remote Sensing Images

The traditional method is used to deal with massive remote sensing data stored in low efficiency and poor scalability. This article presents a parallel processing method based on MapReduce and HBase. The filling of remote sensing images by the Hilbert curve makes the MapReduce method construct pyramids in parallel to reduce network communication between nodes. Then, the authors design a massive remote sensing data storage model composed of metadata storage model, index structure and filter column family. Finally, this article uses MapReduce frameworks to realize pyramid construction, storage and query of remote sensing data. The experimental results show that this method can effectively improve the speed of data writing and querying, and has good scalability.

Large-Scale Remote Sensing Image Retrieval by Deep Hashing Neural Networks

As one of the most challenging tasks of remote sensing big data mining, large-scale remote sensing image retrieval has attracted increasing attention from researchers. Existing large-scale remote sensing image retrieval approaches are generally implemented by using hashing learning methods, which take handcrafted features as inputs and map the high-dimensional feature vector to the low-dimensional binary feature vector to reduce feature-searching complexity levels. As a means of applying the merits of deep learning, this paper proposes a novel large-scale remote sensing image retrieval approach based on deep hashing neural networks (DHNNs). More specifically, DHNNs are composed of deep feature learning neural networks and hashing learning neural networks and can be optimized in an end-to-end manner. Rather than requiring to dedicate expertise and effort to the design of feature descriptors, we can automatically learn good feature extraction operations and feature hashing mapping under the supervision of labeled samples. To broaden the application field, DHNNs are evaluated under two representative remote sensing cases: scarce and sufficient labeled samples. To make up for a lack of labeled samples, DHNNs can be trained via transfer learning for the former case. For the latter case, DHNNs can be trained via supervised learning from scratch with the aid of a vast number of labeled samples. Extensive experiments on one public remote sensing image data set with a limited number of labeled samples and on another public data set with plenty of labeled samples show that the proposed remote sensing image retrieval approach based on DHNNs can remarkably outperform state-of-the-art methods under both of the examined conditions.

Airport Detection Using End-to-End Convolutional Neural Network with Hard Example Mining

Deep convolutional neural network (CNN) achieves outstanding performance in the field of target detection. As one of the most typical targets in remote sensing images (RSIs), airport has attracted increasing attention in recent years. However, the essential challenge for using deep CNN to detect airport is the great imbalance between the number of airports and background examples in large-scale RSIs, which may lead to over-fitting. In this paper, we develop a hard example mining and weight-balanced strategy to construct a novel end-to-end convolutional neural network for airport detection. The initial motivation of the proposed method is that backgrounds contain an overwhelming number of easy examples and a few hard examples. Therefore, we design a hard example mining layer to automatically select hard examples by their losses, and implement a new weight-balanced loss function to optimize CNN. Meanwhile, the cascade design of proposal extraction and object detection in our network releases the constraint on input image size and reduces spurious false positives. Compared with geometric characteristics and low-level manually designed features, the hard example mining based network could extract high-level features, which is more robust for airport detection in complex environment. The proposed method is validated on a multi-scale dataset with complex background collected from Google Earth. The experimental results demonstrate that our proposed method is robust, and superior to the state-of-the-art airport detection models.

Classification of Large-Scale Remote Sensing Images for Automatic Identification of Health Hazards

Remote sensing images from Earth-orbiting satellites are a potentially rich data source for monitoring and cataloguing atmospheric health hazards that cover large geographic regions. A method is proposed for classifying such images into hazard and nonhazard regions using the autologistic regression model, which may be viewed as a spatial extension of logistic regression. The method includes a novel and simple approach to parameter estimation that makes it well suited to handling the large and high-dimensional datasets arising from satellite-borne instruments. The methodology is demonstrated on both simulated images and a real application to the identification of forest fire smoke.

MapReduce-based Parallel Learning for Large-scale Remote Sensing Images

Machine learning applied to large-scale remote sensing images shows inadequacies in computational capability and storage space. To solve this problem, we propose a cloud computing-based scheme for learning remote sensing imag- es in a parallel manner: (1) a hull vector-based hybrid parallel support vector machine model (HHB-PSVM) is proposed. It can substantially improve the efficiency of training and prediction for the large-scale samples while guaranteeing classi- fication accuracy. (2) The MapReduce model is used to achieve parallel extraction of the classification features for the remote sensing images, and the MapReduce-based HHB-PSVM model (MapReduce-HHB-PSVM) is used to implement the training and prediction for large-scale samples. (3) MapReduce-HHB-PSVM is applied to land use classification, ena- bling various types of land use to be classified more efficiently by using fused hyperspectral images. Experimental results show that MapReduce-HHB-PSVM can substantially improve classification efficiency of large-scale remote sensing im- ages while guaranteeing classification accuracy, and it can promote the machine interpretation of ground objects infor- mation extracted from the large-scale remote sensing images to be conducted intelligently.

Convex regularization-based hardware/software co-design for real-time enhancement of remote sensing imagery

In this paper, we address a new approach for high-resolution reconstruction and enhancement of remote sensing (RS) imagery in near-real computational time based on the aggregated hardware/software (HW/SW) co-design paradigm. The software design is aimed at the algorithmic-level decrease of the computational load of the large-scale RS image enhancement tasks via incorporating into the fixed-point iterative reconstruction/enhancement procedures the convex convergence enforcement regularization by constructing the proper projectors onto convex sets (POCS) in the solution domain. The established POCS-regularized iterative techniques are performed separately along the range and azimuth directions over the RS scene frame making an optimal use of the sparseness properties of the employed sensor system modulation format. The hardware design is oriented on employing the Xilinx Field Programmable Gate Array XC4VSX35-10ff668 and performing the image enhancement/reconstruction tasks in a computationally efficient parallel fashion that meets the near-real time imaging system requirements. Finally, we report some simulation results and discuss the implementation performance issues related to enhancement of the real-world RS imagery indicative of the significantly increased performance efficiency gained with the developed approach.