AN EMPIRICAL STUDY ON FULLY CONVOLUTIONAL NETWORK AND HYPERCOLUMN METHODS FOR UAV REMOTE SENSING IMAGERY CLASSIFICATION

Unmanned Aerial Vehicles (UAVs) have revolutionized data acquisition across various domains, presenting immense potential for image processing and semantic segmentation. This literature review encompasses a thorough exploration of advancements, techniques, challenges, and datasets pertaining to UAV image semantic segmentation. It begins by introducing the fundamental concepts of UAVs, highlighting their pivotal role in capturing high-resolution imagery that serves diverse applications. The integration of deep learning algorithms with UAVs is emphasized, unlocking new horizons in autonomous flight, security, and environmental monitoring. Delving into the core principles of semantic segmentation, the review elucidates the critical task of classifying every pixel in an image. Convolutional Neural Networks (CNNs) are presented as the cornerstone technology, tracing their evolution from traditional CNNs to the highly adaptable Fully Convolutional Networks (FCNs). A substantial portion of the review is dedicated to FCNs, underscoring their ability to process images of varying dimensions while maintaining spatial coherence in the output. Their pivotal role in semantic segmentation, encompassing both classification and localization, is articulated. The subsequent sections delve into a comprehensive survey of state-of-the-art models, including SegNet, PSPNet, DeepLabNet, EfficientNet, DenseNet-C, and LinkNet. Each model's unique strengths and applications contribute to the evolving landscape of semantic segmentation tasks. The versatility of the U-Net architecture takes center stage in the latter parts of the review. Its fundamental structure is elucidated, followed by a comprehensive examination of its manifold adaptations—3D-U-Net, ResU-Net, U-Net++, Adversarial U-Net, Cascaded U-Net, and Improved U-Net 3+. These modifications address intrinsic challenges such as limited receptive fields and class imbalances, propelling U-Net to the forefront of image segmentation techniques. The subsequent sections pivot toward the application of U-Net in UAV image segmentation, illustrating its efficacy in diverse tasks, including land cover and crop classification. Nevertheless, persisting challenges, such as the scarcity of annotated datasets and the need for model generalization across varied environmental conditions, remain key areas of concern. The review culminates by underlining the significance of large, authentic datasets and data augmentation techniques. Furthermore, a brief exploration of publicly available UAV image datasets is presented, enhancing our understanding of the resources accessible for training and evaluating models. This comprehensive literature review encapsulates the dynamism of UAV image processing and semantic segmentation, illuminating recent developments and avenues for future research in this burgeoning field.

In the process of Unmanned Aerial Vehicle (UAV) flight testing, plenty of compound faults exist, which could be composed of concurrent single faults or over-limit states alarmed by Built-In-Test (BIT) equipment. At present, there still lacks a suitable automatic labeling approach for UAV flight data, effectively utilizing the information of the BIT record. The performance of the originally employed flight data-driven fault diagnosis models based on machine learning needs to be improved as well. A compound fault labeling and diagnosis method based on actual flight data and the BIT record of the UAV during flight test phase is proposed, through labeling the flight data with compound fault modes corresponding to concurrent single faults recorded by the BIT system, and upgrading the original diagnosis model based on Gradient Boosting Decision Tree (GBDT) and Fully Convolutional Network (FCNN), to eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM) and modified Convolutional Neural Network (CNN). The experimental results based on actual test flight data show that the proposed method could effectively label the flight data and obtain a significant improvement in diagnostic performance, appearing to be practical in the UAV test flight process.

Early detection and treatment are crucial for recovery from cancer. Due to rapid advancement in medical imaging techniques, today doctors can scan various body parts from different perspectives to diagnose cancer or see the progress during a course of treatment. Such techniques are less painful and require less preparation and recovery. The recent evolution of AI, particularly deep learning algorithms have shown potential in computer-aided diagnosis of cancer. These include analysis of cancer images for detection of tumor region, prognosis, skin tissue inspection, etc. Convolutional Neural Network (CNN) based image processing techniques are showing promising results in the diagnosis of breast cancer, skin cancer, prostate cancer, lung cancer, etc. Literature shows many other CNN-based models such as feed-forward neural network, Deep CNN, Multi-Crop CNN, fully convolutional networks (FCNs), Deep Fully Convolutional Network (DFCNet), Recurrent neural networks, etc. The CNN focus solely on the image-specific features and hence require a lesser number of input parameters. A CNN model reduces the size of the input image vector without losing the features critical for making an accurate prediction. This makes it easier to process and reduces computational complexity. CNNs are also not sensitive to the object positioning in an image since the approximate position of image motifs relative to others is known. CNN deals with the variations in images through augmentation and other pre-processing techniques. This makes them robust to various straining conditions in images. This chapter presents an overview of a generic CNN architecture followed by a discussion on major advances in CNN models and their applications for cancer image analysis.

<p>With consumer-grade unmanned aerial vehicles (UAVs) on the rise, which enable easy, time-flexible, and cost-effective acquisition of very high-resolution RGB data, the mapping of forest tree species using solely RGB imagery is of high interest, as it does not rely on sophisticated sensors, does not require extensive calibration and preprocessing and, therefore, enables the application by a wide audience. In combination with convolutional neural networks (CNNs), which particularly exploit spatial patterns and, therefore, highly benfit from very high-resolution remote sensing, this offers great potential for accurately mapping forest tree species.</p><p>Here, we present the findings of our recent study, in which we used very high-resolution RGB imagery from UAVs in combination with CNNs for the mapping of forest tree species. In this study, we used multicopter UAVs to obtain very high-resolution (<2 cm) RGB imagery over 51 ha of temperate forests in the Southern Black Forest region, and the Hainich National Park in Germany. To fully harness the end-to-end learning capabilities of CNNs, we used a semantic segmentation approach (U-net) that concurrently segments and classifies tree species from imagery. With a diverse dataset in terms of study areas, site conditions, illumination properties, and phenology, we accurately mapped nine tree species, three genus-level classes, deadwood, and forest floor (mean F1-score 0.73). We found that a coarser spatial resolution substantially reduced the model accuracy (mean F1-score of 0.26 at 32 cm resolution) and that larger tile sizes during CNN training negatively affected the model accuracies for underrepresented classes. Additional height information from normalized digital surface models slightly increased the model accuracy but simultaneously increased computational complexity and data requirements. Our results highlight the key role that UAVs can play in the mapping of forest tree species, given that air- and spaceborne remote sensing currently does not provide comparable spatial resolutions. Given the end-to-end learning capabilites of CNNs extensive preprocessing becomes partly obsolete, whereas the use of a large and diverse dataset facilitates a high degree of generalization of the CNN, thus fostering transferability. The synergy of high-resolution UAV imagery and CNN provide a fast and flexible yet accurate means of mapping forest tree species.</p><p>In this contribution, we will give an outlook on how the combination of UAV imagery and CNNs can be integrated with multitemporal satellite imagery (Sentinel-1 and Sentinel-2) in order to extrapolate the UAV-based tree species maps to larger areas.</p>

Natural disasters are recurrent weather phenomena whose occurrence has increased worldwide in the past few decades. These disasters cause devastating effects on transportation routes by causing significant damage and obstruction on frequently traveled roads. This research focuses on developing an autonomous network of unmanned aerial vehicles (UAVs) for transportation disaster management using convolutional neural networks (CNNs). The autonomous network of UAVs will allow first responders to optimize their rescue plans by providing relevant information on inaccessible roads. The autonomous UAV system development will increase the affected regions' recovery rate by identifying blocked transportation routes and associating them with their corresponding locations to update the virtual map in real-time. Live footage from the unmanned aerial vehicles is fed to ground control, where the CNN classifies the type of damage encountered and then updates a virtual map through the ArcGIs software. Preliminary results of the classification models such as AlexNet show average accuracy of 74.07%. Furthermore, transfer learning and cross-validation techniques were applied to the CNN models to obtain high confidence levels due to the small dataset size used to train and test the CNNs. To choose the best CNN model, a quantitative analysis was performed to measure the statistical precision, statistical recall, and F1 score on each model to optimize the classification.

The widespread and growing use of flying unmanned aerial vehicles (UAVs) is attributed to their high spatial mobility, autonomous control, and lower cost compared to usual manned flying vehicles. Applications, such as surveying, searching, or scanning the environment with application-specific sensors, have made extensive use of UAVs in fields like agriculture, geography, forestry, and biology. However, due to the large number of applications and types of UAVs, limited power has to be taken into account when designing task-specific software for a target UAV. In particular, the power constraints of smaller UAVs will generally necessitate reducing power consumption by limiting functionality, decreasing their movement radius, or increasing their level of autonomy. Reducing the overhead of control and decision-making software onboard is one approach to increasing the autonomy of UAVs. Specifically, we can make the onboard control software more efficient and focused on specific tasks, which means it will need less computing power than a general-purpose algorithm. In this work, we focus on reducing the size of the computer vision object classification algorithm. We define different tasks by specifying which objects the UAV must recognize, and we construct a convolutional neural network (CNN) for each specific classification. However, rather than creating a custom CNN that requires its dataset, we begin with a pre-trained general-purpose classifier. We then choose specific groups of objects to recognize, and by using response-based pruning (RBP), we simplify the general-purpose CNN to fit our specific needs. We evaluate the pruned models in various scenarios. The results indicate that the evaluated task-specific pruning can reduce the size of the neural model and increase the accuracy of the classification tasks. For small UAVs intended for tasks with reduced visual content, the proposed method solves both the size reduction and individual model training problems.

Unmanned aerial vehicles (UAVs) are invaluable technologies concerning their remote control and monitoring capabilities. Convolutional neural networks (CNNs), known for their high pattern recognition capabilities, are appropriate for forest fire detection with UAVs. Deep convolutional neural networks show substantial performance on hardware with high processing capabilities. While these networks can be operated in unmanned aerial vehicles controlled from ground control stations equipped with GPU-supported hardware, the execution on a typical UAV’s limited computational resources necessitates the use of lightweight, small-sized networks. To overcome these impediments, this article presents a lightweight and attention-based approach for performing forest fire detection tasks using UAV vision data (images acquired by cameras mounted on UAVs). In this paper, we also present comprehensive research for different approaches such as transfer learning, deep CNNs, and lightweight CNNs. Among the experimented models, the attention-based EfficientNetB0 backboned model emerged as the most successful architecture for forest fire detection. With the test accuracy of , the F1-score of , the recall of , and the precision of have strongly reinforced the efficiency of the EfficientNetB0-based model in wildfire recognition. Moreover, the network has a less parameter size than the experimented networks. It proves the model’s suitability for wildfire detection with UAVs having limited hardware resources.

Identification of crops is an important topic in the agricultural domain. Hyperspectral remote sensing data are very useful for crop feature extraction and classification. Remote sensing data is an unstructured data and Convolutional Neural Network (CNN) can work on unstructured data efficiently. This paper presents an evaluation of CNN for crop classification using the Indian Pines standard dataset obtained from the AVIRIS sensor and the study area dataset obtained from the EO-1hyperion sensor. Optimized CNN has been tuned by training the model on different parameters. It has been compared with two classification algorithms: Deep Neural Network (DNN) and Convolutional Autoencoder. According to the test results, the proposed optimized CNN model provided better results as compared to the other two methods. CNN has given 97 ± 0.58% overall accuracy for the Indian Pines standard dataset and 78 ± 2.43% for our study area dataset.

Optical pressure sensors based plantar image segmenting using an improved fully convolutional network

Unmanned aerial vehicles (UAVs) are increasingly being integrated into the domain of precision agriculture, revolutionizing the agricultural landscape. Specifically, UAVs are being used in conjunction with machine learning techniques to solve a variety of complex agricultural problems. This paper provides a careful survey of more than 70 studies that have applied machine learning techniques utilizing UAV imagery to solve agricultural problems. The survey examines the models employed, their applications, and their performance, spanning a wide range of agricultural tasks, including crop classification, crop and weed detection, cropland mapping, and field segmentation. Comparisons are made among supervised, semi-supervised, and unsupervised machine learning approaches, including traditional machine learning classifiers, convolutional neural networks (CNNs), single-stage detectors, two-stage detectors, and transformers. Lastly, future advancements and prospects for UAV utilization in precision agriculture are highlighted and discussed. The general findings of the paper demonstrate that, for simple classification problems, traditional machine learning techniques, CNNs, and transformers can be used, with CNNs being the optimal choice. For segmentation tasks, UNETs are by far the preferred approach. For detection tasks, two-stage detectors delivered the best performance. On the other hand, for dataset augmentation and enhancement, generative adversarial networks (GANs) were the most popular choice.

This study proposes an automatic building footprint extraction framework that consists of a convolutional neural network (CNN)-based segmentation and an empirical polygon regularization that transforms segmentation maps into structured individual building polygons. The framework attempts to replace part of the manual delineation of building footprints that are involved in surveying and mapping field with algorithms. First, we develop a scale robust fully convolutional network (FCN) by introducing multiple scale aggregation of feature pyramids from convolutional layers. Two postprocessing strategies are introduced to refine the segmentation maps from the FCN. The refined segmentation maps are vectorized and polygonized. Then, we propose a polygon regularization algorithm consisting of a coarse and fine adjustment, to translate the initial polygons into structured footprints. Experiments on a large open building data set including 181 000 buildings showed that our algorithm reached a high automation level where at least 50% of individual buildings in the test area could be delineated to replace manual work. Experiments on different data sets demonstrated that our FCN-based segmentation method outperformed several most recent segmentation methods, and our polygon regularization algorithm is robust in challenging situations with different building styles, image resolutions, and even low-quality segmentation.

Understanding the spatial distribution of irrigated croplands is crucial for food security and water use. To map land cover classes with high-spatial-resolution images, it is necessary to analyze the semantic information of target objects in addition to the spectral or spatial–spectral information of local pixels. Deep convolutional neural networks (DCNNs) can characterize the semantic features of objects adaptively. This study uses DCNNs to extract irrigated croplands from Sentinel-2 images in the states of Washington and California in the United States. We integrated the DCNNs of 101 layers, discarded pooling layers, and employed dilation convolution to preserve location information; these are models which were used based on fully convolutional network (FCN) architectures. The findings indicated that irrigated croplands may be effectively detected at various phases of crop growth in the fields. A quantitative analysis of the trained models revealed that the three models in the two states had the lowest values of Intersection over Union (IoU) and Kappa, i.e., 0.88 and 0.91, respectively. The deep models’ temporal portability across different years was acceptable. The lowest values of recall and OA (overall accuracy) from 2018 to 2021 were 0.91 and 0.87, respectively. In Washington, the lowest OA value from 10 to 300 m resolution was 0.76. This study demonstrates the potential of FCNs + DCNNs approaches for mapping irrigated croplands across large regions, providing a solution for irrigation mapping. The spatial resolution portability of deep models could be improved further by designing model architectures.

paper, we focus on exploring the relationship between weave patterns and their mechanical properties in woven fiber composites through Machine Learning. Specifically, we explore the interactions between woven architectures and in-plane stiffness properties through Deep Convolutional Neural Network (DCNN) and Generative Adversarial Network (GAN). Our research is important for exploring how woven composite’s pattern is related to its mechanical properties and accelerating woven composite design as well as optimization. We focus on two tasks: (1) Stiffness prediction: Predicting in-plane stiffness properties for given weave patterns. Our DCNN extracts high-level features through several convolutional and fully connected layers to determine the final predictions. (2) Weave pattern prediction: Predicting weave patterns for target stiffness properties, which can be treated as the reverse task of the first one. Due to many-to-one mapping between weave patterns and the composite properties, we utilize a Decoder Neural Network as our baseline model and compare its performance with GAN and Genetic Algorithm. We represent the weave patterns as 2D checkerboard models and use finite element analysis (FEA) to determine in-plane stiffness properties, which serve as input data for our ML framework. We show that: (1) for stiffness prediction, DCNN can predict stiffness values for a given weave pattern with relatively high accuracy (above 93%); (2) for weave pattern prediction, the GAN model gives the best prediction accuracy (above 92%) while Decoder Neural Network has the best time efficiency. HAOTIAN FENG

The application of deep learning, specifically deep convolutional neural networks (DCNNs), to the classification of remotely-sensed imagery of natural landscapes has the potential to greatly assist in the analysis and interpretation of geomorphic processes. However, the general usefulness of deep learning applied to conventional photographic imagery at a landscape scale is, at yet, largely unproven. If DCNN-based image classification is to gain wider application and acceptance within the geoscience community, demonstrable successes need to be coupled with accessible tools to retrain deep neural networks to discriminate landforms and land uses in landscape imagery. Here, we present an efficient approach to train/apply DCNNs with/on sets of photographic images, using a powerful graphical method called a conditional random field (CRF), to generate DCNN training and testing data using minimal manual supervision. We apply the method to several sets of images of natural landscapes, acquired from satellites, aircraft, unmanned aerial vehicles, and fixed camera installations. We synthesize our findings to examine the general effectiveness of transfer learning to landscape-scale image classification. Finally, we show how DCNN predictions on small regions of images might be used in conjunction with a CRF for highly accurate pixel-level classification of images.

Palm tree detection using remote sensing images has received increasing attention in recent years, concerning the issues of sustainability, productivity and profitability. There has been significant progress in the research of using machine learning techniques, especially convolutional neural networks (CNNs) for automatic palm tree detection. However, whether CNNs can actually outperform traditional human-engineered approaches in terms of classification accuracy and detection speed is yet unknown. In the present study, we have compared human-engineered features namely histogram of oriented gradients (HOG), local binary pattern (LBP) and scale-invariant feature transform (SIFT) with features extracted using pre-trained AlexNet model for detecting palm trees in high resolution images obtained via unmanned aerial vehicle (UAV). Support vector machines (SVM) with linear and non-linear kernels were used to classify feature vectors obtained by different feature extractors. Haar-like features used in Viola-Jones framework was also tested in the study. Results showed that features extracted from the fifth convolutional layer of AlexNet achieved the highest accuracy of 96.1% using SVM with RBF kernel as classifier, which surpassed the accuracy obtained by fully-connected CNN, namely 95.6%. The results also suggest that SVM classifier with radial basis function (RBF) kernel using LBP as features is the optimal combination as it achieved comparable accuracy but higher detection speed than CNN approaches.