Search Results (20)

The study presents a comparative analysis of five distinct image processing methodologies for roadside wild animal detection using thermal imagery, aiming to identify an optimal approach for embedded system implementation to mitigate wildlife–vehicle collisions. The evaluated techniques included the following: bilateral filtering followed by thresholding and SIFT feature matching; Gaussian filtering combined with Canny edge detection and contour analysis; color quantization via the nearest average algorithm followed by contour identification; motion detection based on absolute inter-frame differencing, object dilation, thresholding, and contour comparison; and animal detection based on a YOLOv8n neural network. These algorithms were applied to sequential thermal images captured by a custom roadside surveillance system incorporating a thermal camera and a Raspberry Pi processing unit. Performance evaluation utilized a dataset of consecutive frames, assessing average execution time, sensitivity, specificity, and accuracy. The results revealed performance trade-offs: the motion detection method achieved the highest sensitivity (92.31%) and overall accuracy (87.50%), critical for minimizing missed detections, despite exhibiting the near lowest specificity (66.67%) and a moderate execution time (0.126 s) compared to the fastest bilateral filter approach (0.093 s) and the high-specificity Canny edge method (90.00%). Consequently, considering the paramount importance of detection reliability (sensitivity and accuracy) in this application, the motion-based methodology was selected for further development and implementation within the target embedded system framework. Subsequent testing on diverse datasets validated its general robustness while highlighting potential performance variations depending on dataset characteristics, particularly the duration of animal presence within the monitored frame. Full article

Chagas disease, caused by Trypanosoma cruzi (T. cruzi), remains a significant public health challenge in Latin America. Traditional diagnostic methods relying on manual microscopy suffer from low sensitivity, subjective interpretation, and poor performance in suboptimal conditions. This study presents a novel computer vision framework integrating motion analysis with deep learning for automated T. cruzi detection in microscopic videos. Our motion-based detection pipeline leverages parasite motility as a key discriminative feature, employing frame differencing, morphological processing, and DBSCAN clustering across 23 microscopic videos. This approach effectively addresses limitations of static image analysis in challenging conditions including noisy backgrounds, uneven illumination, and low contrast. From motion-identified regions, $64\times 64$ patches were extracted for classification. MobileNetV2 achieved superior performance with 99.63% accuracy, 100% precision, 99.12% recall, and an AUC-ROC of 1.0. Additionally, YOLOv5 and YOLOv8 models (Nano, Small, Medium variants) were trained on 43 annotated videos, with YOLOv5-Nano and YOLOv8-Nano demonstrating excellent detection capability on unseen test data. This dual-stage framework offers a practical, computationally efficient solution for automated Chagas diagnosis, particularly valuable for resource-constrained laboratories with poor imaging quality. Full article

Accurate cerebral vasculature segmentation using digital subtraction angiography (DSA) is critical for diagnosing and treating cerebrovascular diseases. However, conventional single-frame analysis methods often fail to capture fine vascular structures due to background noise, overlapping anatomy, and dynamic contrast flow. In this study, we propose a novel vessel-enhancing preprocessing technique using temporal differencing of DSA sequences to improve cerebrovascular segmentation accuracy. Our method emphasizes contrast flow dynamics while suppressing static background components by computing absolute differences between sequential DSA frames. The enhanced images were input into state-of-the-art deep learning models, U-Net++ and DeepLabv3+, for vascular segmentation. Quantitative evaluation of the publicly available DIAS dataset demonstrated significant segmentation improvements across multiple metrics, including the Dice Similarity Coefficient (DSC), Intersection over Union (IoU), and Vascular Connectivity (VC). Particularly, DeepLabv3+ with the proposed preprocessing achieved a DSC of 0.83 ± 0.05 and VC of 44.65 ± 0.63, outperforming conventional methods. These results suggest that leveraging temporal information via input enhancement substantially improves small and complex vascular structure extraction. Our approach is computationally efficient, model-agnostic, and clinically applicable for DSA. Full article

Monitoring the activity index of animals is crucial for assessing their welfare and behavior patterns. However, traditional methods for calculating the activity index, such as pixel intensity differencing of entire frames, are found to suffer from significant interference and noise, leading to inaccurate results. These classical approaches also do not support group or individual tracking in a user-friendly way, and no open-access platform exists for non-technical researchers. This study introduces an open-source web-based platform that allows researchers to calculate the activity index from top-view videos by selecting individual or group animals. It integrates Segment Anything Model2 (SAM2), a promptable deep learning segmentation model, to track animals without additional training or annotation. The platform accurately tracked Cobb 500 male broilers from weeks 1 to 7 with a 100% success rate, IoU of 92.21% ± 0.012, precision of 93.87% ± 0.019, recall of 98.15% ± 0.011, and F1 score of 95.94% ± 0.006, based on 1157 chickens. Statistical analysis showed that tracking 80% of birds in week 1, 60% in week 4, and 40% in week 7 was sufficient (r ≥ 0.90; p ≤ 0.048) to represent the group activity in respective ages. This platform offers a practical, accessible solution for activity tracking, supporting animal behavior analytics with minimal effort. Full article

Smoke appears earlier than flames, so image-based fire monitoring techniques mainly focus on the detection of smoke, which is regarded as one of the effective strategies for preventing the spread of initial fires that eventually evolve into serious fires. Smoke monitoring in mine fires faces serious challenges: the underground environment is complex, with smoke and backgrounds being highly integrated and visual features being blurred, which makes it difficult for existing image-based monitoring techniques to meet the actual needs in terms of accuracy and robustness. The conventional ground-based methods are directly used in the underground with a high rate of missed detection and false detection. Aiming at the core problems of mixed target and background information and high boundary uncertainty in smoke images, this paper, inspired by the principle of information entropy, proposes a method for recognizing smoke from mine fires by integrating entropy-enhanced image processing and improved YOLOv8. Firstly, according to the entropy change characteristics of spatio-temporal information brought by smoke diffusion movement, based on spatio-temporal entropy separation, an equidistant frame image differential fusion method is proposed, which effectively suppresses the low entropy background noise, enhances the detail clarity of the high entropy smoke region, and significantly improves the image signal-to-noise ratio. Further, in order to cope with the variable scale and complex texture (high information entropy) of the smoke target, an improvement mechanism based on entropy-constrained feature focusing is introduced on the basis of the YOLOv8m model, so as to more effectively capture and distinguish the rich detailed features and uncertain information of the smoke region, realizing the balanced and accurate detection of large and small smoke targets. The experiments show that the comprehensive performance of the proposed method is significantly better than the baseline model and similar algorithms, and it can meet the demand of real-time detection. Compared with YOLOv9m, YOLOv10n, and YOLOv11n, although there is a decrease in inference speed, the accuracy, recall, average detection accuracy mAP (50), and mAP (50–95) performance metrics are all substantially improved. The precision and robustness of smoke recognition in complex mine scenarios are effectively improved. Full article

Insect detection under real-field conditions remains a challenging task due to factors such as lighting variations and the small size of insects that often lack sufficient visual features for reliable identification by deep learning models. These limitations become especially pronounced in lightweight architectures, which, although efficient, struggle to capture fine-grained details under suboptimal conditions, such as variable lighting conditions, shadows, small object size and occlusion. To address this, we introduce the motion module, a lightweight component designed to enhance object detection by integrating motion information directly at the feature map level within the YOLOv8 backbone. Unlike methods that rely on frame differencing and require additional preprocessing steps, our approach operates on raw input and uses only two consecutive frames. Experimental evaluations demonstrate that incorporating the motion module leads to consistent performance improvements across key metrics. For instance, on the YOLOv8n model, the motion module yields gains of up to 5.11% in mAP50 and 7.83% in Recall, with only a small computational overhead. Moreover, under simulated illumination shifts using HSV transformations, our method exhibits robustness to these variations. These results highlight the potential of the motion module as a practical and effective tool for improving insect detection in dynamic and unpredictable field scenarios. Full article

As the global population ages, robust technological solutions are increasingly necessary to support and enhance elderly autonomy in-home or care settings. This paper presents a novel computer vision-based activity monitoring system that uses cameras and infrared sensors to detect and analyze daily activities of elderly individuals in care environments. The system integrates a frame differencing algorithm with adjustable sensitivity parameters and an anomaly detection model tailored to identify deviations from individual behavior patterns without relying on large volumes of labeled data. The system was validated through real-world deployments across multiple care home rooms, demonstrating significant improvements in emergency response times and ensuring resident privacy through anonymized frame differencing views. Upon detecting anomalies in daily routines, the system promptly alerts caregivers and family members, facilitating immediate intervention. The experimental results confirm the system’s capability for unobtrusive, continuous monitoring, laying a strong foundation for scalable remote elderly care services and enhancing the safety and independence of vulnerable older individuals. Full article

While existing tongue detection methods have achieved good accuracy, the problems of low detection speed and excessive noise in the background area still exist. To address these problems, a fast tongue detection model based on a lightweight model and deep feature propagation (TD-DFP) is proposed. Firstly, a color channel is added to the RGB tongue image to introduce more prominent tongue features. To reduce the computational complexity, keyframes are selected through inter frame differencing, while optical flow maps are used to achieve feature alignment between non-keyframes and keyframes. Secondly, a convolutional neural network with feature pyramid structures is designed to extract multi-scale features, and object detection heads based on depth-wise convolutions are adopted to achieve real-time tongue region detection. In addition, a knowledge distillation module is introduced to improve training performance during the training phase. TD-DFP achieved 82.8% mean average precision (mAP) values and 61.88 frames per second (FPS) values on the tongue dataset. The experimental results indicate that TD-DFP can achieve efficient and accurate tongue detection, achieving real-time tongue detection. Full article

Sensor technologies are increasingly used to monitor laboratory animal behaviour. The aim of this study was to investigate the added value of using accelerometers and video to monitor the activity and drinking behaviour of three rams from 5 days before to 22 days after inoculation with Toxoplasma gondii. We computed the activity from accelerometer data as the vectorial dynamic body acceleration (VDBA). In addition, we assessed individual drinking behaviour from video, using frame differencing above the drinker to identify drinking bouts, and Aruco markers for individual identification. Four days after inoculation, rams developed fever and activity decreased. The daytime VDBA from days 4 to 10 was 60–80% of that before inoculation. Animal caretakers scored rams as lethargic on days 5 and 6 and, for one ram, also on the morning of day 7. Video analysis showed that each ram decreased its number of visits to the drinker, as well as its time spent at the drinker, by up to 50%. The fever and corresponding sickness behaviours lasted until day 10. Overall, while we recognize the limited conclusiveness due to the small number of animals, the sensor technologies provided continuous, individual, detailed, and objective data and offered additional insights as compared to routine observations. We recommend the wider implementation of such technologies in animal disease trials to refine experiments and guarantee the quality of experimental results. Full article

In recent years, there has been a significant increase in satellite launches, resulting in a proliferation of satellites in our near-Earth space environment. This surge has led to a multitude of resident space objects (RSOs). Thus, detecting RSOs is a crucial element of monitoring these objects and plays an important role in preventing collisions between them. Optical images captured from spacecraft and with ground-based telescopes provide valuable information for RSO detection and identification, thereby enhancing space situational awareness (SSA). However, datasets are not publicly available due to their sensitive nature. This scarcity of data has hindered the development of detection algorithms. In this paper, we present annotated RSO images, which constitute an internally curated dataset obtained from a low-resolution wide-field-of-view imager on a stratospheric balloon. In addition, we examine several frame differencing techniques, namely, adjacent frame differencing, median frame differencing, proximity filtering and tracking, and a streak detection method. These algorithms were applied to annotated images to detect RSOs. The proposed algorithms achieved a competitive degree of success with precision scores of 73%, 95%, 95%, and 100% and F1 scores of 68%, 77%, 82%, and 79%. Full article

For staring infrared sensors in geostationary orbit, the clutter caused by the high-frequency jitter and low-frequency drift of the sensor line-of-sight (LOS) is the impact of background features, sensor parameters, LOS motion characteristics, and background suppression algorithms. In this paper, the spectra of LOS jitter caused by cryocoolers and momentum wheels are analyzed, and the time-related factors such as the jitter spectrum, the detector integration time, the frame period, and the temporal differencing background suppression algorithm are considered comprehensively; they are combined into a background-independent jitter-equivalent angle model. A jitter-caused clutter model in the form of multiplying the background radiation intensity gradient statistics by the jitter-equivalent angle is established. This model has good versatility and high efficiency and is suitable for the quantitative evaluation of clutter and the iterative optimization of sensor design. Based on satellite ground vibration experiments and on-orbit measured image sequences, the jitter-caused clutter and drift-caused clutter models are verified. The relative deviation between the model calculation and the actual measurement results is less than 20%. Full article

The wall-to-wall prediction of fuel structural characteristics conducive to high fire severity is essential to provide integrated insights for implementing pre-fire management strategies designed to mitigate the most harmful ecological effects of fire in fire-prone plant communities. Here, we evaluate the potential of high point cloud density LiDAR data from the Portuguese áGiLTerFoRus project to characterize pre-fire surface and canopy fuel structure and predict wildfire severity. The study area corresponds to a pilot LiDAR flight area of around 21,000 ha in central Portugal intersected by a mixed-severity wildfire that occurred one month after the LiDAR survey. Fire severity was assessed through the differenced Normalized Burn Ratio (dNBR) index computed from pre- and post-fire Sentinel-2A Level 2A scenes. In addition to continuous data, fire severity was also categorized (low or high) using appropriate dNBR thresholds for the plant communities in the study area. We computed several metrics related to the pre-fire distribution of surface and canopy fuels strata with a point cloud mean density of 10.9 m⁻². The Random Forest (RF) algorithm was used to evaluate the capacity of the set of pre-fire LiDAR metrics to predict continuous and categorized fire severity. The accuracy of RF regression and classification model for continuous and categorized fire severity data, respectively, was remarkably high (pseudo-R² = 0.57 and overall accuracy = 81%) considering that we only focused on variables related to fuel structure and loading. The pre-fire fuel metrics with the highest contribution to RF models were proxies for horizontal fuel continuity (fractional cover metric) and the distribution of fuel loads and canopy openness up to a 10 m height (density metrics), indicating increased fire severity with higher surface fuel load and higher horizontal and vertical fuel continuity. Results evidence that the technical specifications of LiDAR acquisitions framed within the áGiLTerFoRus project enable accurate fire severity predictions through point cloud data with high density. Full article

With applications such as precise geodetic product generation and reference frame maintenance, the global GNSS network solution is a fundamental problem that has constantly been a focus of concern. Independent baseline search is a prerequisite step of the double-differenced (DD) GNSS network. In this process, only empirical methods are usually used, i.e., the observation-max (OBS-MAX), which allows for obtaining more redundant DD observations, and the shortest-path (SHORTEST), which helps to better eliminate tropospheric and ionospheric errors between stations. Given the possible limitations that neither of the methods can always guarantee baselines of the highest accuracy to be selected, a strategy based on the ‘density’ of common satellites (OBS-DEN) is proposed. It takes the number of co-viewing satellites per unit distance between stations as the criterion. This method ensures that the independent baseline network has both sufficient observations and short baselines. With single-day solutions and annual statistics computed with parallel processing, the method demonstrates that it has the ability to obtain comparable or even higher positioning accuracy than the conventional methods. With a clearer meaning, OBS-DEN can be an option alongside the previous methods in the independent baseline search. Full article

In the past few years, home appliances have been influenced by the latest technologies and changes in consumer trends. One of the most desired gadgets of this time is a universal remote control for gestures. Hand gestures are the best way to control home appliances. This paper presents a novel method of recognizing hand gestures for smart home appliances using imaging sensors. The proposed model is divided into six steps. First, preprocessing is done to de-noise the video frames and resize each frame to a specific dimension. Second, the hand is detected using a single shot detector-based convolution neural network (SSD-CNN) model. Third, landmarks are localized on the hand using the skeleton method. Fourth, features are extracted based on point-based trajectories, frame differencing, orientation histograms, and 3D point clouds. Fifth, features are optimized using fuzzy logic, and last, the H-Hash classifier is used for the classification of hand gestures. The system is tested on two benchmark datasets, namely, the IPN hand dataset and Jester dataset. The recognition accuracy on the IPN hand dataset is 88.46% and on Jester datasets is 87.69%. Users can control their smart home appliances, such as television, radio, air conditioner, and vacuum cleaner, using the proposed system. Full article

A small drone is capable of capturing distant objects at a low cost. In this paper, long distance (up to 1 km) ground target tracking with a small drone is addressed for oblique aerial images, and two novel approaches are developed. First, the coordinates of the image are converted to real-world based on the angular field of view, tilt angle, and altitude of the camera. Through the image-to-position conversion, the threshold of the actual object size and the center position of the detected object in real-world coordinates are obtained. Second, the track-to-track association is improved by adopting the nearest neighbor association rule to select the fittest track among multiple tracks in a dense track environment. Moving object detection consists of frame-to-frame subtraction and thresholding, morphological operation, and false alarm removal based on object size and shape properties. Tracks are initialized by differencing between the two nearest points in consecutive frames. The measurement statistically nearest to the state prediction updates the target’s state. With the improved track-to-track association, the fittest track is selected in the track validation region, and the direction of the displacement vector and velocity vectors of the two tracks are tested with an angular threshold. In the experiment, a drone hovered at an altitude of 400 m capturing video for about 10 s. The camera was tilted 30° downward from the horizontal. Total track life (TTL) and mean track life (MTL) were obtained for 86 targets within approximately 1 km of the drone. The interacting multiple mode (IMM)-CV and IMM-CA schemes were adopted with varying angular thresholds. The average TTL and MTL were obtained as 84.9–91.0% and 65.6–78.2%, respectively. The number of missing targets was 3–5; the average TTL and MTL were 89.2–94.3% and 69.7–81.0% excluding the missing targets. Full article