Search Results (211)

Marine species photo-identification and location for tracking are crucial for understanding the characteristics and patterns that distinguish each marine species. However, challenges in camera data acquisition and the unpredictability of animal movements have restricted progress in this field. To address these challenges, we present a novel algorithm for the first stage of marine species photo-identification and location methods. For marine species photo-identification applications, a color index-based thresholding segmentation method is proposed. This method is based on the characteristics of the GMR (Green Minus Red) color index and the proposed empirical BMG (Blue Minus Green) color index. These color indexes are modified to provide better information about the color of regions, such as marine animals, the sky, and land found in the scientific sightings images, allowing an optimal thresholding segmentation method. In the case of marine species location, a SURFs (Speeded-Up Robust Features)-based supervised classifier is used to obtain the location of the marine animal in the sighting image; with this, its tracking could be obtained. The tests were performed with the Kaggle happywhale public database; the results obtained in precision shown range from 0.77 up to 0.98 using the proposed indexes. Finally, the proposed method could be used in real-time marine species tracking with a processing time of 0.33 s for images of 645 × 376 pixels using a standard PC. Full article

The threat posed by unauthorized drones to public airspace has become increasingly critical. To address the challenge of UAV detection in unaligned visible–infrared dual-spectral images, we present a novel framework that comprises two sequential stages: image alignment and object detection. The Speeded-Up Robust Features (SURF) algorithm is applied for feature matching, combined with the gray centroid method to remove mismatched feature points. A plane-adaptive pixel remapping algorithm is further developed to achieve images fusion. In addition, an enhanced YOLOv11 model with a modified loss function is employed to achieve robust object detection in the fused images. Experimental results demonstrate that the proposed method enables precise pixel-level dual-spectrum fusion and reliable UAV detection under diverse and complex conditions. Full article

Registering UAV-based thermal and visible images is a challenging task due to differences in appearance across spectra and the lack of public benchmarks. To address this issue, we introduce UAV-TIRVis, a dataset consisting of 80 accurately and manually registered UAV-based thermal (640 × 512) and visible (4K) image pairs, captured across diverse environments. We benchmark our dataset using well-known registration methods, including feature-based (ORB, SURF, SIFT, KAZE), correlation-based, and intensity-based methods, as well as a custom, heuristic intensity-based method. We evaluate the performance of these methods using four metrics: RMSE, PSNR, SSIM, and NCC, averaged per scenario and across the entire dataset. The results show that conventional methods often fail to generalize across scenes, yielding <0.6 NCC on average, whereas the heuristic method shows that it is possible to achieve 0.77 SSIM and 0.82 NCC, highlighting the difficulty of cross-spectral UAV alignment and the need for further research to improve optimization in existing registration methods. Full article

Road agglomerate fog seriously threatens driving safety, making real-time fog state detection crucial for implementing reliable traffic control measures. With advantages in aerial perspective and a broad field of view, UAVs have emerged as a novel solution for road agglomerate fog monitoring. This paper proposes an agglomerate fog detection method based on the fusion of SURF and optical flow characteristics. To synthesize an adequate agglomerate fog sample set, a novel network named FogGAN is presented by injecting physical cues into the generator using a limited number of field-collected fog images. Taking the region of interest (ROI) for agglomerate fog detection in the UAV image as the basic unit, SURF is employed to describe static texture features, while optical flow is employed to capture frame-to-frame motion characteristics, and a multi-feature fusion approach based on Bayesian theory is subsequently introduced. Experimental results demonstrate the effectiveness of FogGAN for its capability to generate a more realistic dataset of agglomerate fog sample images. Furthermore, the proposed SURF and optical flow fusion method performs higher precision, recall, and F1-score for UAV perspective images compared with XGBoost-based and survey-informed fusion methods. Full article

In patients with respiratory failure, mechanical ventilation aims to balance respiratory muscle loading and gas exchange. The interplay between the ventilator and the respiratory muscles is an increasingly recognized factor in tailoring ventilatory support. Surface electromyography (sEMG) offers a non-invasive modality to monitor the respiratory muscles. The sEMG signal, however, requires elaborate processing, which is limitedly standardized and documented. This paper presents the Respiratory Surface Electromyography (ReSurfEMG) package, an open-source Python package for respiratory sEMG analysis developed to address these challenges. ReSurfEMG integrates denoising, feature extraction, and quality assessment in one dedicated library. The effects of over- and under-filtering were compared to ReSurfEMG default settings regarding waveform duration, time-to-peak, amplitude, electrical time product (ETP), pseudo-slope, pseudo-signal-to-noise ratio (SNR), area under the baseline (AUB), and bell-curve error. Under-filtering increased amplitudes (+21%) and ETPs (+10%). Over-filtering smoothed sEMG waveforms, reducing amplitude (−58%), ETP (−39%), and pseudo-slope (−49%), while waveform duration and time-to-peak increased. Default ReSurfEMG settings provided the highest SNRs with similar or lower AUBs and bell-curve errors. The ReSurfEMG library integrates advanced methods dedicated to respiratory sEMG analysis. Systematic assessment using ReSurfEMG showed that signal processing settings affect sEMG features. ReSurfEMG enables reproducible signal processing, facilitating the standardization of respiratory sEMG analysis. Full article

To address the challenge of feature extraction and reconstruction in weak-texture environments, and to provide data support for environmental perception in mobile robots operating in such environments, a Feature Extraction and Reconstruction in Weak-Texture Environments solution is proposed. The solution enhances environmental features through laser-assisted marking and employs a two-step feature extraction strategy in conjunction with binocular vision. First, an improved SURF algorithm for feature point fast localization method (FLM) based on multi-constraints is proposed to quickly locate the initial positions of feature points. Then, the robust correction method (RCM) for feature points based on light strip grayscale consistency is proposed to calibrate and obtain the precise positions of the feature points. Finally, a sparse 3D (three-dimensional) point cloud is generated through feature matching and reconstruction. At a working distance of 1 m, the spatial modeling achieves an accuracy of ±0.5 mm, a relative error of 2‰, and an effective extraction rate exceeding 97%. While ensuring both efficiency and accuracy, the solution demonstrates strong robustness against interference. It effectively supports robots in performing tasks such as precise positioning, object grasping, and posture adjustment in dynamic, weak-texture environments. Full article

This study introduces a novel adaptive Mechanical Wave Energy Converter (MWEC) designed to efficiently capture nearshore wave energy for sustainable electricity generation along the southeast surf coast of NEOM (135° longitude). The MWEC system features a polyvinyl chloride (PVC) cubic buoy integrated with a mechanical power take-off (PTO) mechanism, optimized for deployment in shallow waters for a depth of around 1 m. Three buoy volumes, V1: 6000 cm³, V2: 30,000 cm³, and V3: 72,000 cm³, were experimentally evaluated under consistent PTO and spring tension configurations. The findings reveal a direct relationship between buoy volume and force output, with larger buoys exhibiting greater energy capture potential, while smaller buoys provided faster and more stable response dynamics. The energy retention efficiency of the buoy–PTO system was measured at 20% for V1, 14% for V2, and 10% for V3, indicating a trade-off between responsiveness and total energy capture. Notably, the largest buoy (V3) generated a peak power output of 213 W at an average wave amplitude of 65 cm, confirming its suitability for high-energy conditions along NEOM’s surf coast. In contrast, the smaller buoy (V1) performed more effectively during periods of reduced wave activity. Wave climate data collected during November and December 2024 support a hybrid deployment strategy, utilizing different buoy sizes to adapt to seasonal wave variability. These results highlight the potential of modular, wave-adaptive mechanical systems for scalable, site-specific renewable energy solutions in coastal environments like NEOM. The proposed MWEC offers a promising path toward low-cost, low-maintenance wave energy harvesting in shallow waters, contributing to Saudi Arabia’s sustainable energy goals. Full article

Image matching plays a critical role in a wide range of computer vision applications, including object recognition, 3D reconstruction, aiming-point and six-degree-of-freedom detection for aiming devices, and video surveillance. Over the past three decades, image-matching algorithms and techniques have evolved significantly, from handcrafted feature extraction algorithms to modern approaches powered by deep learning neural networks and attention mechanisms. This paper provides a comprehensive review of image-matching techniques, aiming to offer researchers valuable insights into the evolving landscape of this field. It traces the historical development of feature-based methods and examines the transition to neural network-based approaches that leverage large-scale data and learned representations. Additionally, this paper discusses the current state of the field, highlighting key algorithms, benchmarks, and real-world applications. Furthermore, this study introduces some recent contributions to this area and outlines promising directions for future research, including H-matrix optimization, LoFTR model speedup, and performance improvements. It also identifies persistent challenges such as robustness to viewpoint and illumination changes, scalability, and matching under extreme conditions. Finally, this paper summarizes future trends for research and development in this field. Full article

Face recognition technology utilizes unique facial features to analyze and compare individuals for identification and verification purposes. This technology is crucial for several reasons, such as improving security and authentication, effectively verifying identities, providing personalized user experiences, and automating various operations, including attendance monitoring, access management, and law enforcement activities. In this paper, comprehensive evaluations are conducted using different face detection and modality segmentation methods, feature extraction methods, and classifiers to improve system performance. As for face detection, four methods are proposed: OpenCV’s Haar Cascade classifier, Dlib’s HOG + SVM frontal face detector, Dlib’s CNN face detector, and Mediapipe’s face detector. Additionally, two types of feature extraction techniques are proposed: hand-crafted features (traditional methods: global local features) and deep learning features. Three global features were extracted, Scale-Invariant Feature Transform (SIFT), Speeded Robust Features (SURF), and Global Image Structure (GIST). Likewise, the following local feature methods are utilized: Local Binary Pattern (LBP), Weber local descriptor (WLD), and Histogram of Oriented Gradients (HOG). On the other hand, the deep learning-based features fall into two categories: convolutional neural networks (CNNs), including VGG16, VGG19, and VGG-Face, and Siamese neural networks (SNNs), which generate face embeddings. For classification, three methods are employed: Support Vector Machine (SVM), a one-class SVM variant, and Multilayer Perceptron (MLP). The system is evaluated on three datasets: in-house, Labelled Faces in the Wild (LFW), and the Pins dataset (sourced from Pinterest) providing comprehensive benchmark comparisons for facial recognition research. The best performance accuracy for the proposed ten-feature extraction methods applied to the in-house database in the context of the facial recognition task achieved 99.8% accuracy by using the VGG16 model combined with the SVM classifier. Full article

To address the high-precision navigation requirements of urban low-altitude electric vertical take-off and landing (eVTOL) aircraft in environments where global navigation satellite systems (GNSSs) are denied and under complex urban terrain conditions, a terrain-matching optimization algorithm based on light detection and ranging (LiDAR) is proposed. Given the issues of GNSS signal susceptibility to occlusion and interference in urban low-altitude environments, as well as the error accumulation in inertial navigation systems (INSs), this algorithm leverages LiDAR point cloud data to assist in constructing a digital elevation model (DEM). A terrain-matching optimization algorithm is then designed, incorporating enhanced feature description for key regions and an adaptive random sample consensus (RANSAC)-based misalignment detection mechanism. This approach enables efficient and robust terrain feature matching and dynamic correction of INS positioning errors. The simulation results demonstrate that the proposed algorithm achieves a positioning accuracy better than 2 m in complex scenarios such as typical urban canyons, representing a significant improvement of 25.0% and 31.4% compared to the traditional SIFT-RANSAC and SURF-RANSAC methods, respectively. It also elevates the feature matching accuracy rate to 90.4%; meanwhile, at a 95% confidence level, the proposed method significantly increases the localization success rate to 96.8%, substantially enhancing the navigation and localization accuracy and robustness of eVTOLs in complex low-altitude environments. Full article

A key component in visual Simultaneous Location And Mapping (SLAM) systems is feature extraction and description. One common algorithm that accomplishes this purpose is Oriented FAST and Rotated BRIEF (ORB), which is used in state-of-the-art SLAM systems like ORB-SLAM. While it is faster than other feature detectors like SIFT (340 times faster) or SURF (15 times faster), it is one of the most computationally expensive algorithms in these types of systems. This problem has commonly been solved by delegating this task to hardware-accelerated solutions like FPGAs or ASICs. While this solution is useful, it incurs a greater economical cost. This work proposes a solution for feature extraction and description based on a modern low-power mainstream microcontroller. The execution time of ORB, along with power consumption, are analyzed in relation to the number of feature points and internal variables. The results show a maximum of 0.6 s for ORB execution in 1241 × 376 resolution images, which is significantly slower than other hardware-accelerated solutions but remains viable for certain applications. Additionally, the power consumption ranges between 30 and 40 milliwatts, which is lower than FPGA solutions. This work also allows for future optimizations that will improve the results of this paper. Full article

Face anti-spoofing detection is crucial for identity verification and security monitoring. However, existing single-modal models struggle with feature extraction under complex lighting conditions and background variations. Moreover, the feature distributions of live and spoofed samples often overlap, resulting in suboptimal classification performance. To address these issues, we propose a jointly optimized framework integrating the Enhanced Channel Attention (ECA) mechanism and the Intra-Class Differentiator (ICD). The ECA module extracts features through deep convolution, while the Bottleneck Reconstruction Module (BRM) employs a channel compression–expansion mechanism to refine spatial feature selection. Furthermore, the channel attention mechanism enhances key channel representation. Meanwhile, the ICD mechanism enforces intra-class compactness and inter-class separability, optimizing feature distribution both within and across classes, thereby improving feature learning and generalization performance. Experimental results show that our framework achieves average classification error rates (ACERs) of 2.45%, 1.16%, 1.74%, and 2.17% on the CASIA-SURF, CASIA-SURF CeFA, CASIA-FASD, and OULU-NPU datasets, outperforming existing methods. Full article

Accurate acquisition of two-dimensional digital maps of disaster sites is crucial for rapid and effective emergency response. The construction of two-dimensional digital maps using unmanned aerial vehicle (UAV) aerial images is not affected by factors such as signal interference, terrain, or complex building structures, which are common issues with methods like single-soldier image transmission or satellite imagery. Therefore, this paper investigates a method for modeling two-dimensional digital maps based on UAV aerial images. The proposed Canny edge-enhanced Speeded-Up Robust Features (C-SURF) algorithm in this method is designed to enhance the number of feature extractions and the accuracy of image registration. Compared to the SIFT and SURF algorithms, the number of feature points increased by approximately 44%, and the registration accuracy improved by about 16%, laying a solid foundation for feature-based image stitching. Additionally, a novel image stitching method based on the novel energy function is introduced, effectively addressing issues such as color discrepancies, ghosting, and misalignment in the fused image sequences. Experimental results demonstrate that the signal-to-noise ratio (SNR) of the fused images based on the novel energy function can reach an average of 36 dB. Full article

Because single local or global characteristics can only depict the classification information of an object unilaterally or partially, that may result in low recognition accuracy; in this paper we propose an improved SURF and modified Zernike moments descriptor (ISMZMD) for object recognition. Firstly, we extracted the improved SURF and seven modified Zernike moments descriptors of objects. Secondly, we effectively fused the two features together with different weight factors based on their contribution to object identification. Thirdly, we computed the Euclidean distance to decide the recognition result. Finally, we evaluated the performance of the proposed algorithm and compared it with other algorithms. The results of the experiments show that our algorithm is effective and robust to scaling alteration, translation change, rotation variation, and noise transformation. Compared with other representative methods, our method has a higher recognition rate and less recognition time. Full article

This paper introduces a hybrid prediction method that combines the Bi-LSTM neural network with definitions of surf-riding, wave-blocking and broaching to enhance the safety and stability of ship navigation. The hybrid method can accurately predict ships’ attitude motions and states by recognizing ship states and encoding them into one-hot representations. The Bi-LSTM model’s bidirectional learning capability captures significant temporal dependencies, enabling precise and timely predictions of complex maritime events across various conditions. Additionally, the direct output approach of state features improves prediction accuracy by eliminating intermediate steps, allowing for the better anticipation of and response to critical events. Validated with ship navigation data from autopilot simulations in wave conditions, the hybrid method outperforms conventional methods based on LSTM and Bi-LSTM models, demonstrating strong generalization capabilities and significantly contributing to safer and more stable ship navigation. Full article