Objects In Images Research Articles

Improved Re-Parameterized Convolution for Wildlife Detection in Neighboring Regions of Southwest China.

To autonomously detect wildlife images captured by camera traps on a platform with limited resources and address challenges such as filtering out photos without optimal objects, as well as classifying and localizing species in photos with objects, we introduce a specialized wildlife object detector tailored for camera traps. This detector is developed using a dataset acquired by the Saola Working Group (SWG) through camera traps deployed in Vietnam and Laos. Utilizing the YOLOv6-N object detection algorithm as its foundation, the detector is enhanced by a tailored optimizer for improved model performance. We deliberately introduce asymmetric convolutional branches to enhance the feature characterization capability of the Backbone network. Additionally, we streamline the Neck and use CIoU loss to improve detection performance. For quantitative deployment, we refine the RepOptimizer to train a pure VGG-style network. Experimental results demonstrate that our proposed method empowers the model to achieve an 88.3% detection accuracy on the wildlife dataset in this paper. This accuracy is 3.1% higher than YOLOv6-N, and surpasses YOLOv7-T and YOLOv8-N by 5.5% and 2.8%, respectively. The model consistently maintains its detection performance even after quantization to the INT8 precision, achieving an inference speed of only 6.15 ms for a single image on the NVIDIA Jetson Xavier NX device. The improvements we introduce excel in tasks related to wildlife image recognition and object localization captured by camera traps, providing practical solutions to enhance wildlife monitoring and facilitate efficient data acquisition. Our current work represents a significant stride toward a fully automated animal observation system in real-time in-field applications.

Exploring machine vision measurements through innovative cyclic-lead-follower statistical technique: an experimental study

In metrology systems, machine vision systems are often utilized for non-contact inspection. The most important phase in ensuring measurement accuracy is camera calibration and estimation of pixel measurement errors, which establish the correspondence between image coordinates and object coordinates. Multiple calibration techniques improve the effectiveness of machine vision systems. However, a number of factors lead to variations in the camera calibration procedure, which must be optimized. This study explains a novel ‘Cyclic-Lead-Follower’ statistical methodology proposed for camera calibration and measurement to estimate the errors in pixel measurement, employing four slip gauges for measurement. Several multi-criteria decision-making techniques, including WSM, WPM, WASPAS, and TOPSIS, were used to optimize the results of the proposed Cyclic-Lead-Follower methods. The proposed Cyclic-Lead-Follower method improves the accuracy of the camera calibration and measurement system, which employs the exponential moving average statistical method when compared to the traditional calibration method. The proposed calibration method produces lower exponential moving average values than the traditional calibration method, with an average percentage error of approximately 46% in the exponential moving average. The use of an exponential moving average in a validation experiment of the Cyclic-Lead-Follower method decreased the measurement percentage errors, which were estimated to be 0.57%. The proposed method can be used in machine vision systems due to its robustness, accuracy, and cost-effectiveness.

Deep learning techniques to detect rail indications from ultrasonic data for automated rail monitoring and maintenance

The increasing number of passengers and services using railways and the corresponding increase in rail use has caused the acceleration of rail wear and surface defects which makes rail defect identification an important issue for rail maintenance and monitoring to ensure safe and efficient operation. Traditional visual inspection methods for identifying rail defects are time-consuming, less accurate, and associated with human errors. Deep learning has been used to improve railway maintenance and monitoring tasks. This study aims to develop a structured model for detecting railway artifacts and defects by comparing different deep-learning models using ultrasonic image data. This research showed whether it is practical to identify rail indications using image classification and object detection techniques from ultrasonic data and which model performs better among the above-mentioned methods. The methodology includes data processing, labeling, and using different conventional neural networks to develop the model for both image classification and object detection. The results of CNNs for image classification, and YOLOv5 for object detection show 98%, and 99% accuracy respectively. These models can identify rail artifacts efficiently and accurately in real-life scenarios, which can improve automated railway infrastructure monitoring and maintenance.

AFDet: alignment and focusing for aerial object detection

ABSTRACT Object detection in aerial images has become a focus in recent years due to the expansion of its application fields. Aerial objects have the characteristics of multi-scale, arbitrary angle and dense arrangement, which brings considerable challenges to the task. Based on the efficient one-stage detectors, most methods design modules to adaptively align the feature receptive field with the anchors to improve the accuracy of the bounding box regression and calculate loss according to parameter bias. Current methods are not adequately data-driven in feature optimization and the angle parameter regression faces boundary problems. To handle these problems, in this letter, we propose an Align-Focus detector (AFDet) for aerial image object detection. We introduce the ideology of deformable multi-head self-attention to feature optimization and design a new Align-Focus Module (AFM) which can sensitively focus feature encoding response to the real texture area of the objects. In addition, we apply KF-IOU Loss to solve the boundary problem. We conduct necessary experiments on DOTA and HRSC2016 datasets, and AFDet achieves the highest mAP performance compared to the existing methods.

Gamma camera imaging characteristics of 166Ho and 99mTc used in Selective Internal Radiation Therapy

BackgroundThe administration of a 166Ho scout dose is available as an alternative to 99mTc particles for pre-treatment imaging in Selective Internal Radiation Therapy (SIRT). It has been reported that the 166Ho scout dose may be more accurate for the prediction of microsphere distribution and the associated therapy planning. The aim of the current study is to compare the scintigraphic imaging characteristics of both isotopes, considering the objectives of the pre-treatment imaging using clinically geared phantoms.MethodsPlanar and SPECT/CT images were obtained using a NEMA image quality phantom in different phantom setups and another body-shaped phantom with several inserts. The influence of collimator type, count statistics, dead time effects, isotope properties and patient obesity on spatial resolution, contrast recovery and the detectability of small activity accumulations was investigated. Furthermore, the effects of the imaging characteristics on personalized dosimetry are discussed.ResultsThe images with 99mTc showed up to 3 mm better spatial resolution, up to two times higher contrast recovery and significantly lower image noise than those with 166Ho. The contrast-to-noise ratio was up to five times higher for 99mTc than for 166Ho. Only when using 99mTc all activity-filled spheres could be distinguished from the activity-filled background. The measurements mimicking an obese patient resulted in a degraded image quality for both isotopes.ConclusionsOur measurements demonstrate better scintigraphic imaging properties for 99mTc compared to 166Ho in terms of spatial resolution, contrast recovery, image noise, and lesion detectability. While the 166Ho scout dose promises better prediction of the microsphere distribution, it is important to consider the inferior imaging characteristics of 166Ho, which may affect individualized treatment planning in SIRT.

Optimizing DNN training with pipeline model parallelism for enhanced performance in embedded systems

Deep Neural Networks (DNNs) have gained widespread popularity in different domain applications due to their dominant performance. Despite the prevalence of massively parallel multi-core processor architectures, adopting large DNN models in embedded systems remains challenging, as most embedded applications are designed with single-core processors in mind. This limits DNN adoption in embedded systems due to inefficient leveraging of model parallelization and workload partitioning. Prior solutions attempt to address these challenges using data and model parallelism. However, they lack in finding optimal DNN model partitions and distributing them efficiently to achieve improved performance.This paper proposes a DNN model parallelism framework to accelerate model training by finding the optimal number of model partitions and resource provisions. The proposed framework combines data and model parallelism techniques to optimize the parallel processing of DNNs for embedded applications. In addition, it implements the pipeline execution of the partitioned models and integrates a task controller to manage the computing resources. The experimental results for image object detection demonstrate the applicability of our proposed framework in estimating the latest execution time and reducing overall model training time by almost 44.87% compared to the baseline AlexNet convolutional neural network (CNN) model.

Automatic Marker Generation to Similarity Based Region Merging Algorithm using Edge Information

One of the most important tasks in computer vision is image segmentation. Several interactive strategies are utilized for picture segmentation because automated techniques are difficult for this kind of work. The outcome of interactive methods is mostly determined by user input. Obtaining excellent interactions for huge datasets is challenging. However, automated picture segmentation is starting to play a significant role in image analysis and computer vision. Effective and efficient segmentation outcomes are obtained using the interacting region merging method suggested by Maximal Similarity based region merging algorithm. Limitation of MSBRM is it requires some efforts on the part of users and is yet not a fully-automatic approach. Here, we suggest a completely fresh unsupervised image segmentation method that combines edge information with MSBRM. We propose an integrated framework to generate object markers for similarity based region algorithm using edge information. Long edges give rough distribution of objects in image. After retrieving edges using phase congruency, edge processing operations are employed to remove small edges and to group color similar long boundaries. Centroids of long boundaries are used as object markers to the MSBRM algorithm. The generation of object markers is done using edge segment grouping. These object markers guide the region merging process. The proposed method shows its effectiveness in segmenting natural real world color images.

DIGCN: A Dynamic Interaction Graph Convolutional Network Based on Learnable Proposals for Object Detection

We propose a Dynamic Interaction Graph Convolutional Network (DIGCN), an image object detection method based on learnable proposals and GCN. Existing object detection methods usually work on dense candidates, resulting in redundant and near-duplicate results. Meanwhile, non-maximum suppression post-processing operations are required to eliminate negative effects, which increases the computational complexity. Although the existing sparse detector avoids cumbersome post-processing operations, it ignores the potential relationship between objects and proposals, which hinders detection accuracy improvement. Therefore, we propose a dynamic interaction GCN module in the DIGCN, which performs dynamic interaction and relational modeling on the proposal boxes and proposal features to improve the object detection accuracy. In addition, we introduce a learnable proposal method with a sparse set of learned object proposals to eliminate a huge number of hand-designed object candidates, avoiding complicated tasks such as object candidate design and many-to-one label assignment, and reducing object detection model complexity to a certain extent. DIGCN demonstrates accuracy and run-time performance on par with the well-established and highly optimized detector baselines on the challenging COCO dataset, e.g. with the ResNet-101FPN as the backbone our method attains the accuracy of 46.5 AP while processing 13 frames per second. Our work provides a new method for object detection research.

Digital brand image management based on feedback loops

Purpose: is to propose a model for managing a digital brand image brand based on consecutive feedback loops from the image targetaudience. Methods: collection, processing, and analysis of secondary data were carried out on the basis of desk research, synthesis methods, contentanalysis, and systemic, structural and functional data evaluation using an interdisciplinary approach. Results: systemized methodological approaches in Russian and foreign practice by the authors served as the basis for development of a digital brand image management model based on consecutive feedback loops from the image target audience. Within the framework of the proposed model, it is supposed to create systems in a digital environment for providing feedback to the image target audience and monitoring of formed relationships with consumers. The authors' methodological recommendations for implementation of the proposed model are manifested in detailed approaches to organizing the relationship of the image object with the image target audience through different types of feedback loops, methods of monitoring responses and reactions of target consumers, methods of assessing image in a brand's digital environment to close feedback loops. Conclusions and Relevance: loops of feedback from the brand’s image target audience are a causative mechanism of dynamic nature in which the output data of subjective and objective consumer assessment of brand reputation is used as input data for the management and enhancement of the digital image. Setting up a system of feedback from consumers through digital channels on the basis of consecutive loops allows you to correlate planned indicators of the brand image with the actual indicators, as well as identifi cation of the existing problems and introducing timely adjustments to decisions on managing a brand’s digital image.

Application of Object Contrast for Forming Images Used in Technical Vision Systems for Navigation of Mobile Robots

The purpose of the article is to substantiate the feasibility of using object contrast as an informative feature for the formation of images used in technical vision systems. This goal is achieved by studying the dependence of the contrast of objects in images on the viewing geometry and determining the conditions under which the greatest similarity of the compared images is ensured. The solution to the first problem is based on the presentation of reference information about sighted objects, taking into account the navigation parameters of mobile robots. By modeling in the MATLAB software environment for typical viewing conditions, selective images were obtained using randomly selected fragments from Google Earth Pro, the distribution of contrast values and the cross-correlation function of the original and selective images. The influence of viewing angles on the distribution of contrasts and the formation of the decisive function was determined. The studies were performed for viewing angles of -60˚, -80˚ and -90˚ for altitudes in the range from 500 to 600 meters. The most significant result is a model of a set of reference images, taking into account the influence of navigation parameters on the contrast of objects, as well as experimentally established dependences of the distribution of contrasts for typical viewing conditions. The novelty of the work lies in the fact that the procedure for generating images and the decision function using the contrast of objects has been further developed. This will significantly increase the efficiency of selection of objects with insignificant brightness characteristics.

A Convolutional Neural Network for Tropical Cyclone Wind Structure Identification in Kilometer-Scale Forecasts

Abstract Detection and tracking of tropical cyclones (TCs) in numerical weather prediction model outputs is essential for many applications, such as forecast guidance and real-time monitoring of events. While this task has been automated in the 1990s with heuristic models, relying on a set of empirical rules and thresholds, the recent success of machine learning methods to detect objects in images opens new perspectives. This paper introduces and evaluates the capacity of a convolutional neural network based on the U-Net architecture to detect the TC wind structure, including maximum wind speed area and hurricane-force wind speed area, in the outputs of the convective-scale AROME model. A dataset of 400 AROME forecasts over the West Indies domain has been entirely hand-labeled by experts, following a rigorous process to reduce heterogeneities. The U-Net performs well on a wide variety of TC intensities and shapes, with an average intersection-over-union metric of around 0.8. Its performances, however, strongly depend on the TC strength, and the detection of weak cyclones is more challenging since their structure is less well defined. The U-Net also significantly outperforms an operational heuristic detection model, with a significant gain for weak TCs, while running much faster. In the last part, the capacity of the U-Net to generalize on slightly different data is demonstrated in the context of a domain change and a resolution increase. In both cases, the pretrained U-Net achieves similar performances as the original dataset.

Application of convolutional neural networks trained on optical images for object detection in radar images

Due to the small number of annotated radar image datasets, the use of optical images for training neural networks designed to detect objects in radar images seems promising. However, optical images have some significant differences from radar images and an experimental investigation of this possibility is required. In this work we investigate the applicability of such an approach and show that in the case of detection of ships good results can be achieved. In addition, it is shown that preliminary filtering of speckle noise can improve the results.

Contrastive Masked Autoencoders are Stronger Vision Learners.

Masked image modeling (MIM) has achieved promising results on various vision tasks. However, the limited discriminability of learned representation manifests there is still plenty to go for making a stronger vision learner. Towards this goal, we propose Contrastive Masked Autoencoders (CMAE), a new self-supervised pre-training method for learning more comprehensive and capable vision representations. By elaboratively unifying contrastive learning (CL) and masked image model (MIM) through novel designs, CMAE leverages their respective advantages and learns representations with both strong instance discriminability and local perceptibility. Specifically, CMAE consists of two branches where the online branch is an asymmetric encoder-decoder and the momentum branch is a momentum updated encoder. During training, the online encoder reconstructs original images from latent representations of masked images to learn holistic features. The momentum encoder, fed with the full images, enhances the feature discriminability via contrastive learning with its online counterpart. To make CL compatible with MIM, CMAE introduces two new components, i.e., pixel shifting for generating plausible positive views and feature decoder for complementing features of contrastive pairs. Thanks to these novel designs, CMAE effectively improves the representation quality and transfer performance over its MIM counterpart. CMAE achieves the state-of-the-art performance on highly competitive benchmarks of image classification, semantic segmentation and object detection. Notably, CMAE-Base achieves 85.3% top-1 accuracy on ImageNet and 52.5% mIoU on ADE20k, surpassing previous best results by 0.7% and 1.8% respectively.

Ultimate pose estimation: A comparative study

AbstractPose estimation is a computer vision task used to detect and estimate the pose of a person or an object in images or videos. It has some challenges that can leverage advances in computer vision research and others that require efficient solutions. In this paper, we provide a preliminary review of the state‐of‐the‐art in pose estimation, including both traditional and deep learning approaches. Also, we implement and compare the performance of Hand Pose Estimation (HandPE), which uses PoseNet architecture for hand sign problems, for an ASL dataset by using different optimizers based on 10 common evaluation metrics on different datasets. Also, we discuss some related future research directions in the field of pose estimation and explore new architectures for pose estimation types. After applying the PoseNet model, the experiment results showed that the accuracy achieved was 99.9%, 89%, 97%, 79%, and 99% for the ASL alphabet, HARPET, Yoga, Animal, and Head datasets, comparing those with common optimizers and evaluation metrics on different dataset.

Review of Poultry Monitoring using Computer Vision

Poultry farming is an important unit in global agronomy, contributing immensely to the production of meat and eggs. Safeguarding the health and welfare of poultry is vital for ethical and financial reasons. In recent years, computer vision awareness has gained prominence as a powerful tool for poultry monitoring. This review paper provides an outline of the application of computer vision in poultry monitoring. We explore the different phases of this technology, with real-time image acquisition, object recognition, and behavior analysis. By connecting cameras and sophisticated algorithms, a computer vision system can distinguish unusual behavior, track poultry activities and observe its environmental conditions. Furthermore, this study discovers the benefits of computer vision in poultry farming, including early detection of disease, better production efficiency and improved animal welfare. In conclusion, the application of computer vision in monitoring poultry holds immense potential for the industry; by providing a corridor to a more sustainable and ethical poultry farming system with increased productivity. This paper equally discusses recent developments, challenges, forthcoming prospects in the field and academic research gaps identified.

IMAGE PROCESSING APPROACH FOR FOREIGN MATERIAL DETECTION IN COTTON BUNDLE

The image processing philosophy is mainly determined by the complexity of the image and provides the necessary information to be derived from the image. In the textile industry, the image processing technique focuses on the determination of the geometric properties of the fibers such as cross-sectional shape, diameter, length, fineness, and curl while the studies on the yarn characteristics mostly focus on the determination of yarn hairiness, yarn unevenness and yarn defects (thick place, thin place and neps). In this study, previous studies about image processing approaches that are applied for fiber characteristics were investigated. A case study was conducted to automatically determine the visible foreign matter in the waste cotton bundle that can be used for recycled cotton yarn production. It was revealed that the image processing methods can be successfully applied for foreign fiber and matter detection in cotton bundle. As a result, it is emphasized that the waste cotton properties can be specified with a sensitive and accurate approach via image processing technique, objective and numerical determination can be obtained instead of visual evaluation based on experience.

Hyperparameter Optimization Using Hyperband in Convolutional Neural Network for Image Classification of Indonesian Snacks

Abstract. Indonesia is known for its traditional food both domestically and abroad. Several cakes are included in favorite traditional foods. Of the many types of cakes that exist, it is visually easy to recognize by humans, but computer vision requires special techniques in identifying image objects to types of cakes. Therefore, to recognize objects in the form of images of cakes as one of Indonesian specialties, a deep learning algorithm technique, namely the Convolutional Neural Network (CNN) can be used. Purpose: This study aims to find out how the Convolutional Neural Network (CNN) works by optimizing the hyperband hyperparameter in the classification process and knowing the accuracy value when hyperband is applied to the optimal hyperparameter selection process for classifying Indonesian snack images. Methods/Study design/approach: This study optimizes the hyperparameter Convolutional Neural Network (CNN) using Hyperband on the Indonesian cake dataset. The dataset is 1845 images of Indonesian snacks which consists of 1523 training data, 162 validation data and 160 testing data with 8 classes. In training data, the dataset is divided by 82% on training data, 9% validation, and 9% testing. Result/Findings: The best hyperparameter value produced is 480 for the number of dense neurons 2 and 0.0001 for the learning rate. The proposed method succeeded in achieving a training value of 87.53%, for the validation process it was obtained 66.8%, the testing process was obtained 79.37%. Results obtained from model training of 50 epochs. Novelty/Originality/Value: Previous research focused on the application and development of algorithms for the classification of Indonesian snacks. Therefore, optimizing hyperparameters in a Convolutional Neural Network (CNN) using Hyperband can be an alternative in selecting the optimal architecture and hyperparameters.

TENSOR AND VECTOR APPROACHES TO OBJECTS RECOGNITION BY INVERSE FEATURE FILTERS

The investigation of the extraction of image objects features by filters based on tensor and vector data presentation is considered. The tensor data is obtained as a sum of rank-one tensors, given by the tensor product of the vector of lexicographic representation of image fragments pixels with itself. The accumulated tensor is approximated by one rank tensor obtained using singular values decomposition. It has been shown that the main vector of the decomposition can be considered as the object feature vector. The vector data is obtained by accumulating analogous vectors of image fragments pixels. The accumulated vector is also considered as an object feature. The filter banks of a set of objects are obtained by regularized inversion of the matrices compiled by object features vectors. Optimized regularization of the inversion is used to expand the regions of object features capture with minimal error. The object fragments and corresponding feature vectors are selected through a training iterative process. The tensor and vector approaches create two channels for recognition. High efficiency of object recognition can be achieved by choosing the filter capture band and creating filter branches according to the given bands. The filters create a convolutional network to recognize a set of objects. It has been shown that the obtained filters have an advantage over known correlation filters when recognizing objects with small fragments.

A lightweight multi-feature fusion network for unmanned aerial vehicle infrared ray image object detection

UAV (Unmanned Aerial Vehicle) infrared object detection is crucial in pedestrian monitoring and traffic dispatch, which detects and locates objects in infrared images. In light of issues such as unnoticeable texture features and limited resolution of infrared image objects, a lightweight multi-scale feature fusion method for UAV infrared object detection is presented to enhance the performance of UAVs carrying intelligent devices to detect infrared objects. By changing the anchorless frame strategy of the YOLOX method, a lightweight Multi-Feature Fusion Network (MFFNet) for UAV infrared ray (IR) image object detection is proposed. First, a lightweight backbone network is built using ShuffleNetv2_block, spatial pyramid pooling, and other modules to reduce the network's number of parameters and inference time while maintaining its capacity to extract features. Second, we develop a multi-feature fusion module to improve the detection capabilities of the model for IR objects by fusing the local features and the overall characteristics of IR objects since the texture features of IR objects are challenging to employ, but the boundary information is evident. The boundary frame regression loss is then optimized using SCYLLA-IoU (SIoU) by comparing the predicted frame to the actual frame in terms of angle, distance, shape, and IoU (Intersection over Union), which forces the model to reach the optimum predicted box more quickly. The experimental results demonstrate that our method achieves an 81.5% mean average precision (mAP) with 4.21M parameters and an inference time of only 4.84ms per image, outperforming most networks in speed and accuracy.

An investigation into deep learning for the analysis of medical images

In recent years, deep learning has emerged as a pivotal paradigm in the analysis of medical images, with convolutional networks serving as a cornerstone of this advancement. This paper delves into a comprehensive exploration of the fundamental principles underpinning deep learning and its applications within the domain of medical image analysis. Through a meticulous review of many contemporary contributions, this study synthesizes the latest developments in the field, emphasizing tasks like image classification, object detection, segmentation, and registration. The inquiry spans diverse medical disciplines, encompassing neurology, retinal imaging, pulmonary studies, digital pathology, breast and cardiac evaluations, and musculoskeletal analyses. As a culmination, the paper not only assesses the present state-of-the-art achievements but also critically discusses persistent challenges and illuminates promising avenues for future research endeavors.